diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 6.288937943981838,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.8441,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.109546302948798,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.8262,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.4351071351697025,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.8693,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.163289759641453,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.8301,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 5.951247690464702,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.8188,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 5.95537733759356,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.8281,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.56994495888051,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.7754,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.384598931978227,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.7966,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.4340921061743193,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.7499,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.1951982741427316,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.7229,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.142892757604305,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.7432,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.7034501970705627,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.7156,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.329261093644419,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.7236,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.039599025375866,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.7106,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.9177585815694767,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.6835,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.5348693067277526,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.6895,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.073925759842721,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.6478,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.548154698445663,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.6315,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.7899239315666502,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.6083,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3544913522142057,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.6047,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.5181109790267706,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.5848,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 2.1077346701443633,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.6116,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 2.1070335316856443,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.6058,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.5228444148290539,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.602,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.8618080665784698,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.5828,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1182519020064605,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.5128,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.3282740730519822,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.5592,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0345151068870306,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.552,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.6906341532013363,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.5741,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8819249197359316,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.5423,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9248369488462284,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.5399,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.7640380325118066,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.5426,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.6710000351541291,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.532,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.7882045059397359,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.5302,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.790491045628924,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.5525,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.5871300826084545,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.5228,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7011567826212585,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.5308,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6727378179014453,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.5308,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.629400520372807,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.5222,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6205858040394512,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.4966,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7868127236471973,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5171,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.6356855316092364,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.5096,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.6635055200540139,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.5169,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6268446731610499,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.5123,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6227747160229375,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.5205,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6832333100331771,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.5103,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.5946616856378039,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.4943,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6521908477681193,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.4712,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6098018373595379,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.5283,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.6300289535706199,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.4784,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6831660198942077,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.5119,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.45856544775554964,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.4749,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6905976591643509,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.4942,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.5794840325307908,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.5004,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6851011781737316,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.4847,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7020546004676742,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.4762,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.6545458456113422,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.4891,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8481193918682235,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.5009,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.6131364458421756,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.4787,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.7069734680695235,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.4939,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.6679242988265006,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.4884,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.7377089078256925,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.4987,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.7182982227174018,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4467,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.6534635473262469,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4374,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6592732992596322,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.4456,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.6237489108725235,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4541,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.6664422296030025,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.4893,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.827714789855873,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.4953,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.6249357550511558,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.4629,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7212448647524291,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.4509,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.6363152860132528,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.4663,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7349806849501918,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.4995,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.8561726767890975,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.4907,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.0888013985843832,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.4701,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.5455686010619334,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.4584,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.7513590580152251,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.4546,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8938759542826801,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.4844,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.0259227046619295,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.4743,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1378110458613946,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.489,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.1217933631469696,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.504,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8703969403943289,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.4652,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.3675551716878607,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4717,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.087339320215252,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.4588,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.2427776511962276,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.4608,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0628594387235775,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.4675,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.4286969491720913,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.4644,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.1502878535862326,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.4398,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.4461354785865566,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.4671,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.1690157675881878,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4729,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.231152559632141,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.4689,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.885070551554189,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.464,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.5188594732566807,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.4738,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.1188493521985152,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.4647,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.3228245429376395,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.4755,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.245521902853895,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.4822,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.2236080092862625,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.4688,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.0286621214521254,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.4492,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.3615460110956685,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.4649,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.1524155064910162,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.467,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.093715877404018,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.4408,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.2064180308059234,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.4451,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.161067185122057,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.4638,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.189967913998219,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.4412,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.7767681726177239,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.4475,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.4353958859326683,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.4542,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.8447660930180665,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.4544,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.9738891244420931,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.4538,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.946215682507166,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.4403,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.0360854664996888,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.4649,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8175474165563027,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.4339,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.9069473610751286,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.4541,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9132608188068253,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.4415,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.1341975400124897,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.4251,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.1566100991790618,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.4726,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.0057357043503654,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.455,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.271560397447077,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.4796,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.081123549648899,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.4663,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.1599597258268899,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4564,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.029138286580922,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.4895,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.3736173699773466,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.4457,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.9797366184138858,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.4439,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.7545059238598857,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4662,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.4160962517807547,
+      "learning_rate": 4e-05,
+      "loss": 0.4495,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.4661614310548925,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.4484,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.311108008797089,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.4359,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.5463944666969593,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.4558,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.6502923637951297,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.4548,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0935229809543328,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.4717,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.6547442324304218,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.4335,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.1278963157441955,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.4639,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.571538637139211,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.438,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.463862562373148,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.4664,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.3256275174780858,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.4641,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.4860382974164403,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.445,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.2050945566624236,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.4479,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.6855566741048997,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.4614,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.081020148063881,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.4566,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.3387052081027813,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.4547,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.9837544276013784,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.461,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.392236310615117,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.4298,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.099699664542618,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.4292,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.1454219733227287,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.4657,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.9208034348173921,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.4427,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.023959450688765,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.4425,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.8977725511215772,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.4399,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.7960632358180002,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.467,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.8564303085691354,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.4474,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8703101421753041,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.4185,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.9132874708232781,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.4488,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.730066890302058,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.4296,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7463920913199132,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.4452,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8782067382852146,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.4615,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7684273005935497,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.4524,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.6824850330507772,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.4679,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.008584904625179,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.4685,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.1219704549264096,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.4428,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.8005081858459168,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.4496,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7197621908533207,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.4714,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.7972411426019578,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.4498,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.131114282801615,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.4579,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.7202527339462124,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.4405,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.8284264496501936,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.4543,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.067603686145978,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.4568,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.880345786553939,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.4357,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.7409500000314846,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.4565,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.7890240931500141,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.4494,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.663116283227692,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.4192,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6022250729746302,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.4388,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.633764676133467,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.4276,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.7114841810641654,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.4489,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.6537862529359941,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.4697,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.733556085813622,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.422,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7813141596335805,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.4799,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.84348683742878,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.4608,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7617409741632949,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.4393,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.7821457664215223,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.4431,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.9198610657886641,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.4418,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.811373102903043,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.4499,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.7244948043756048,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.4555,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.8329539287636304,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.4613,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.7946474848890898,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.4306,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.7578050175565608,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.4558,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.9197138428417017,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.4411,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.58657608221511,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.4165,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6991232208483331,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.4246,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6489334044691883,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.4414,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.7081498135076908,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.4115,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.7127785996263198,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.4352,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.916382629708847,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.448,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.0148345966816315,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.4271,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.7001622192433256,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.4212,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7521122315332641,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.4452,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.5881708043350192,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.4148,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6697655755882191,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.4443,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5714814433672581,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.4307,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.7003420391094366,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.4588,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.582674905404985,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.4051,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6079140376153734,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.4442,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.5250093746946322,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.4476,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6068754326562812,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.4015,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5597760583057377,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.4263,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5630330229233381,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.4359,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6688410178997041,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.4509,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5734248261595657,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.4141,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6459897904911214,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.3919,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.615625384223484,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.4293,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.6723709518473275,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.4165,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.7456948813229997,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.4438,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.6167038007835721,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.4109,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.7518596094201674,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.4399,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.8390356269933013,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.4211,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.7312521723080837,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.419,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6096270327955043,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.4512,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.9453469241431455,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.4202,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6976237086060217,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.4284,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.7556596727439536,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.4392,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6676926166367659,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.418,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7104027489524956,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.4288,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6470881776757944,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.4136,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.7467935998236508,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.4218,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5393336482824115,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.4567,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.7399540085659314,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.4155,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5602364486052526,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.4275,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.8459595038085324,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.4305,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.5336385353502059,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.4231,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7179740131205208,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.4243,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6049845916036599,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.4348,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.5907730564292211,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.43,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6162282240699893,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.4226,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.5398551415838272,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.4259,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6948824680487732,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.4032,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.7585193458405427,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.4213,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6348856654646673,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.4166,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.583854935550841,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.4383,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.758572063452047,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.4112,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6725473728216966,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.4209,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5870501872661845,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.4219,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6745106909402518,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.4294,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6626603467066815,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.4378,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.7068984175452087,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.4102,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.7880421001654906,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.4331,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.7980939005440107,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.4293,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6876872116520303,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.4119,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.693340437136292,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.4242,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6445294497640396,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.4096,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6204537564924009,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4311,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.9337164486786037,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.4794,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.7453446177722538,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.361,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.5684176352246474,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.3553,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.5700376389064562,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.3401,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.6963369598283751,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.3756,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.8338734047190438,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.3869,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.718808531788869,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.3544,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.636318653412764,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.3927,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.8540121225067054,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.3629,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.7157538689575355,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.3482,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7150977332758838,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.3751,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8812622357065288,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3433,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.7587297620782933,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.3854,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7796136768002714,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.372,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.5527805649125126,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.3778,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.8684429656193673,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.3642,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.475915575302659,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.3355,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.7150516911355861,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.3406,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.6089289810371407,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.36,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.609949465599699,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.3801,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6435792174595134,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.3692,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.5371366134818112,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.3742,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.5581120409138158,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.3699,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.6871539859922031,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.3847,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.6728625389506454,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.3522,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.5798052022631193,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.3471,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.6554344004927534,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.3862,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.616773525655812,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.3514,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6252073032907827,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.3625,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.7292203014835472,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.3743,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6750961802769572,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.3562,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.641266563707821,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.3811,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.47626082346617665,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.3439,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.775345466369956,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.3949,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.6512639870587921,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.3425,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.508138088788479,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.3743,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.8417082858421004,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.3715,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.6150245525170723,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.3555,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.5368056397944975,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.3711,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.8245806878213414,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.3828,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6332679783321343,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.3756,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.5919233946878811,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.3693,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.9003372181846105,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.3333,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.7225673064353961,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.3563,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.6253600689054303,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.3647,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.7487247600339301,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.3842,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.6729261090984905,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.372,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.5711718821845027,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.3572,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.6269092835925634,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.3638,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.7227144586966063,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.3456,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.5168130646422068,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.3444,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.6534190556974251,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.3981,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.49608264060985185,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.3528,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.5873973891310795,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.3547,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.5190849418755,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.3769,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.6580351021516313,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.3785,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.5007015612418153,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.328,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.6002877300819153,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.3534,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.534656525368796,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.3723,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.6475123076622619,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.4,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.4996754620184398,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.3446,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.6044314095618284,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.3463,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6701066680782941,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.3816,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6601978195740922,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.3636,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.654031675054211,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.3549,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.8149362156444989,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.3717,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.5252172782917475,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.3647,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.7097815767038591,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.3971,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6013979505793712,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.3509,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5372447231288704,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.3454,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.4819156870780376,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.3657,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.5532517256347196,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.3754,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.5730453231624307,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.3362,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.47986747991599304,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.3545,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.6192302663851932,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.3564,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.5999883261505492,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.385,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.5245158098906796,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.3592,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.6103943930252037,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.3448,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.6059126444916411,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.3534,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.59094251729358,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.3462,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.5116820988760775,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.3601,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.6104912515576283,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3715,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5843502579002078,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3615,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.57278755941287,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.379,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5845426558356979,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.3622,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.6031293753243854,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.3516,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.7601747680949098,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.385,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.611379409373343,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.3795,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.769914973216462,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.3598,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.5108484284473499,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.3696,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.6603533182175088,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.3703,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.6378298600911905,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.3711,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.5746667327110595,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.3873,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5877808941315068,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.3373,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.44508675884944343,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.3699,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6514338703924268,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.3616,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.488084612420718,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.3885,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.48826909107605676,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.3551,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.5165747841765426,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.358,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.5512632702594142,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.3764,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.554886404479317,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.3697,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.5571208330365793,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.3521,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.557322603006468,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.3583,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6535997143676556,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.3731,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.5601328122132363,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.3503,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6043062851656329,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.3669,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.49147266214738433,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.3584,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.4208053083010604,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.3268,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.5504152483410247,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.3885,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.44171954199553987,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3546,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5223205552879934,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.357,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.47608725088434434,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.3449,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.6369980078864992,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.3886,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.45042098087237586,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.3677,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.7524250746468709,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.3501,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.5913483123117511,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3838,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.7096543009814014,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.3754,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.5941330284322239,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.3502,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6082956702247818,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.3442,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.5410578343185274,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.3826,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.5603622049356614,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.3543,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6354714740868278,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.3673,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.5482904707182389,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3706,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.7373400164876783,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3778,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.557750313024634,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.3811,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6619066192689937,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.3439,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.5328887963392257,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.3467,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.5176856044780699,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.3631,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6667778969706615,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3845,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.46997105453610205,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.3673,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.5660540216490277,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.3528,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.5232169837718346,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.3927,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.5301059775758672,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.3343,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.48978127332685245,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.3743,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.5974101904334694,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.3873,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.4610781527988556,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.3356,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5476193802797952,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.3678,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.5681899640534382,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.3793,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.46579117952622723,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.3692,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.45829967350822226,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.3861,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.5312267538767446,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.3567,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.4927270455560705,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.3837,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.6730254520540553,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3708,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.49271123224050245,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.3607,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6015749958083195,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.3666,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5331756287325641,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.3465,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5461383098432756,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.3625,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.46173693948299266,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.3447,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4338080042120724,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.3665,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.6364201364886273,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.3973,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.38780557160498685,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.3628,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.65437279914286,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.3696,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.45658993546964116,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.3541,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6341167117991405,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.3552,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.534677010166471,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.344,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6469146129161396,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.4067,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6486336400614103,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3339,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.4756165734311412,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.3656,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.47986132720212116,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.362,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5282839538434738,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.3898,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.42198371123997747,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.3853,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5492741811221751,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.327,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.4942059343945595,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.3569,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6624639978821428,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.353,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.5054436143179456,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.3739,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.5578084759961504,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.3479,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5175138230312964,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3788,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.4767316272043443,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.3659,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.4615644095466124,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.3495,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.48121159480802045,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.3544,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.43389797724079665,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.3345,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5422182984952244,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.3945,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.4346395054624208,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.3722,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.5171227379353726,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.3282,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.47023784691089826,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.3741,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.712171914938611,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.3762,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.4614979661851374,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.3552,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.7029034559542094,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.3765,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.5107503396271503,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.3718,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.6484696790744422,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.3416,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.5422611956095955,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.3701,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.5736593916941034,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.381,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.516655441055286,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3382,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.5810986477370371,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.381,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.4965966913528199,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.3541,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.5253427734627145,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.3618,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.5731585348702032,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3399,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.5650586511844738,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.3614,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.5552795332526816,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.338,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.4676224005337027,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.3388,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.5761255035413918,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.3728,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.4376577074505231,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3422,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.6202523482194091,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.3805,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.5657236383416656,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.3723,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5765939002069779,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.3403,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5126622978326395,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.3846,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.4893918662887372,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.346,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.49480945200066695,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.3608,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.46371077578922587,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.3691,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.5177247078034167,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.3453,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.5395717808296161,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.3854,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.48328591418691574,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.332,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.5120844496620333,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.3318,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.4986397845114079,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.3704,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.471067460408827,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.3259,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.5781762554004196,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.3773,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5284883200306638,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.3608,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.5747925576788419,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.3523,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.48333748512532687,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3738,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.4936158863072384,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.385,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.4730438485920699,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.3475,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.44666350886468653,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.3633,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6610011686297157,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.3809,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.539347729488059,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.3687,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.7734304824540277,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3537,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.5555651572753354,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.3609,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.47054719858651245,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.345,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.46389859286482593,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.3686,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.48643754713664666,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.366,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5004078196211961,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.3508,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.41998158304865213,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.3691,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.6057146754888291,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.3753,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.41195974586516804,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.3704,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.5358781822764654,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.3483,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.42371008689948536,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.344,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.5346562750014451,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.3856,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.4141720915009854,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3806,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5172239795185752,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3598,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.4511973588972147,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.3681,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5092352114684312,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.3519,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5570310032445711,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.3765,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.47311189360012185,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.3491,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.47056397565238867,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.3427,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.4695186144871229,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.3725,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.422241233020545,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3452,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.5183444581588461,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.3744,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.4890424471615685,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3444,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.4717431186408318,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.362,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5085937929000877,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.4045,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.42791490906171814,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.3268,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.43123646544264344,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3411,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.4436697196802756,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3557,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.4908985503356589,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.3706,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.42805255034421297,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.3392,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.4761690732867276,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.3893,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5379733974383426,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3366,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.5032533768082184,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.3474,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.5417903317815617,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3791,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.6016681098915937,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.2869,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.5821715415970093,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2773,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.6574321516027156,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.2805,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.6097083485815958,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.2654,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.5808259327045548,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.292,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.6147385328374464,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2527,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5244169580443896,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.2817,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6190881521829624,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.2573,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.4940972315357483,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.2597,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.5145668087207806,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2743,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.5237924256472535,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.298,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.4913436567569768,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2633,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.47735221640844244,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2812,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.44336247120712663,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.2658,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.4024700019811851,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.266,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.4162880782814302,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.2739,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.420900234997548,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.2587,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.3922810256675891,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.2684,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.44039350319646303,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2817,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.39407208755129564,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.266,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.4870609706168049,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.2742,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.38689750819994556,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2848,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.47724906144643825,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.2599,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.4401164469574589,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3054,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.4053578101644078,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.2759,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.40305386272038873,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.2773,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.43650044464884125,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.2874,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.3478101167004297,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.238,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5378540131732767,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2819,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.36452648854664715,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.2664,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5032510399114508,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.2953,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.39046265531624147,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2853,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.43727290072759506,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.2667,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.3512656906091559,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.2433,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.4269997892462966,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.272,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.40210985884153905,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.281,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.45494580927402933,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.2823,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.3786584829429648,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.2901,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.4158341663125679,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.2503,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.4973773297094156,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3136,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.358332971278896,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.25,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5402190477927518,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.2934,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.41553091019951377,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2405,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.36953009494537353,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.2702,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.4618430786249228,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.2754,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.34300639365902447,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2661,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.42270755166468227,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3022,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.45565654557923757,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.2829,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.3740877825201809,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.2543,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.41128932845003974,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.2599,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.403860224442484,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2797,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.4361999470271837,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3052,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.3532067760107498,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2347,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.4440493361283224,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.2995,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.3502379795099289,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.2609,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.4012490759619669,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.263,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.38804170877650906,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.2827,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.39588544055041436,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.2845,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3708984978870487,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2627,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.43399891803241697,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2726,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.41283712831926583,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.2793,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.497039098012783,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.2585,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.3786496548955617,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.2604,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.41087510783106246,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2619,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.3862921393907714,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2755,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.402183622318764,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.2821,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.41359965199052373,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2689,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.4057314734351054,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.2818,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.4380612193658628,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3026,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.3872729608545391,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2676,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.37305161844563806,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.2682,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.42294171630040434,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2926,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.32895047708719866,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.2495,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.41679444610604394,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3011,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.3679230855847808,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.2504,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4423043701268338,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2766,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.3677471887988386,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.274,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.40692499333061316,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.2778,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.3263572188311098,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.2523,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.4222393366651242,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.2887,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.40331562296694845,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.2867,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.4013137908626882,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.2841,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.40860920576967896,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.2361,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.4763124990319127,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.297,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.42322085204991683,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.281,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.44776968378610954,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.2949,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.37435167719085816,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.241,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.513191206195513,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2864,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.40346995376463735,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.2637,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.4494500627672261,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.2803,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.42726884777191765,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.2767,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.41346524264591156,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.2676,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.4341772184028144,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2777,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4099315160781371,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2561,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.5138255737584566,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3042,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.432761183615062,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.2651,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.46443280746712196,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2832,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.4164805297513381,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.2805,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.4779495850904071,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2938,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5019425790082251,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3005,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.41441111529602026,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.2681,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.39810232392019645,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.2647,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.45088722260930886,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.2792,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4433210427214778,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2934,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.4066976478814104,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2918,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.34370967093720184,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.2544,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.3968350071149765,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3179,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.3397607968212906,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.2445,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.3761869224783527,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2739,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.4107165943497234,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3034,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.38702317295513144,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2887,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.36784735360345644,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2782,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.3336713985884654,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.254,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.3584980258444885,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2777,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.33734236236087706,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.2709,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.3698143393700194,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2925,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.40817337309250085,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3037,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.37767872425813503,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.2681,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.3676604700971711,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.2693,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.35333735391132376,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.2416,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.44264024806603935,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2977,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.35863912278809695,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.2673,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.38072283234786136,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.2412,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.403597150881794,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3105,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.3782480685217854,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2777,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.3605196714297788,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.2475,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.3551162684454082,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2992,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.3932563953219981,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.2695,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.36364661597582887,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.2809,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.40023260576552105,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.2769,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.37722053190398297,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.2651,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.3609591056261206,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.269,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.37366824482919936,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.2751,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.384625652392506,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2907,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.3698678730179469,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.2603,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 3.249664228310474,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2982,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.460430005261284,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.2615,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.4118932913498316,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.266,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.434244203097639,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.263,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.41319653394567957,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.2738,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.4440300391307443,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2806,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.4450317588914836,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2749,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4157434440791862,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.2741,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5170206500260366,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2814,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.36857928753881886,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.2749,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.49606621721335087,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.2844,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.39734223538593466,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2902,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.49898987544665974,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2621,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.4801726281957128,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.2947,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.4065929670620292,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.2453,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.49073219506877014,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3074,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.4297273904385644,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.26,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.39347970454754255,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.2838,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.39976746077544734,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2591,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.43302153534298704,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.2751,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.3303177896338,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.2614,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 9.399730562979176,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.2817,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.5992193214680664,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.2881,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.3495970220376663,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.2663,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 115.53762551096254,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3395,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.9651412902709733,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.2772,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4495442114831174,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.2924,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.7845739846504417,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2711,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.604608692044039,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.2661,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.6258294439759592,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.263,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5494913594798536,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.2794,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5558555891434741,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.2839,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.515162337281719,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.2745,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5344311712053955,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2884,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.42018336642438764,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.2589,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.41996023591809356,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2827,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.4438210925909113,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.2965,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.46076402925545595,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2961,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.43000908691025647,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.2713,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.39043927792582855,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.2718,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.46931796941276205,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.286,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.37000567362431025,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2597,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.3679585756339379,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.268,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.39390644187051665,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2942,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.35844080148715257,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.2856,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.3571063496253924,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.277,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.31448535357403046,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2893,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.41094958660060343,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3032,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.33881077645371754,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.2442,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.40486201913248177,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.2774,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.40045559615117493,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2915,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4049476085253368,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.2917,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.4021405560783678,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.259,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.484899516021424,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2802,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.36800834466998755,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.31,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.38673479130020283,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.2377,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.4332943511576832,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.2649,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.36300549033307267,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3031,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.4305837053016307,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.2828,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.3502213335789738,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.2785,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.4119184464066906,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.2852,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.3328116936278123,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.2451,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.40104590315772765,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2705,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.35485351647827795,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.2787,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.30002141404598737,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.2499,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.3925946973591372,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.2893,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.3768155023282516,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2872,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.3545992718800668,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2846,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.36785601865583023,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.283,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.33508488153952887,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.2521,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.38302973066847995,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2685,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.36992938702258926,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.2807,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.3862905492854937,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2946,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.40907160443716484,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.266,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4109122050114267,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.2802,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.354127837344062,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.2721,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.3870972705130291,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.2631,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5433003414723407,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2784,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.3737974816735393,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.2631,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.3649373556343311,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2611,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.45692117298430235,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.306,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.37167317870118693,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.2604,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.418570087880976,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.2805,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.40393401191819966,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.2942,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.35302339242416125,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.2734,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.31056291236856537,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.2555,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.4273996521286828,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3075,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.33943055418412144,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.2695,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.33579307893991234,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.2632,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.3851698016450561,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.2568,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.3429084587191431,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.288,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.3630187622420414,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.2568,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.40613240314017124,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.28,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.3630057912609447,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.2921,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.38114810144733857,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2902,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.3758839839656435,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.2665,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.3595084530365976,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2706,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.35572078789652456,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.2656,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.433237523934633,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2919,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.3484099523558317,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.277,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.3758907827631485,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.2672,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.37842937900538176,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2841,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3608536152659806,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.2993,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.331034786951403,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.2459,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.381139816381753,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3112,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.32683414482125434,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2663,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.36576519429132925,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.2739,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.374985512809441,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.2865,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.3265013082221716,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2571,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.44682764239559847,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3052,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.341825063992544,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.2829,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.45149506764990327,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2702,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.4725091122888919,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1974,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.36621281865362865,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2001,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.4181656353691187,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1604,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.6522357664764024,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2045,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.3967649572732406,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.187,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.4368447770797467,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1828,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.38796594935236806,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.1848,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.38219443790672963,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2049,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.3555089092501948,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1897,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.3394913053973038,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1689,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.3734166328747449,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1893,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.38780348674547066,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1983,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 11.610555554661177,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2677,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.500081382546598,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2096,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.35588647792299377,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.191,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.37065966241562853,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1834,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.38889585495492435,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1939,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.3837358900844906,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2086,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.3201444685205819,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1696,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4083670291658414,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2022,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.3038168242937323,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1748,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.3667487135723708,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1957,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.3099372569798142,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1583,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.36804096468793185,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1952,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.32919329386761337,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1895,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.3281715049317034,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1869,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.34393105563460474,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1738,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.30224685746831453,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1925,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.31531154058119065,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1818,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.35576844142406727,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1958,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.321407868710026,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1889,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.323403830811808,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1805,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.3258698669038706,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1844,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.3468257128548358,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1954,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.27102202471964076,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1687,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.3391057810178958,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1922,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 11.993272984185353,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2552,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.35342298553688967,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1921,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.30535979200463514,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1849,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.3298566020941755,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.196,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.30677379836364155,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1788,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.29172597513958254,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1804,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.3467487818294283,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1935,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.3453347664050692,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1922,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.27441192409347875,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1625,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.33076055827351086,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.1921,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.3143474634273232,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1936,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3531619059917717,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1866,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.3450353414704144,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2019,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.32436547179369907,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2106,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.3348905345178557,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1851,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.338011127028556,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.174,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.3217953703310517,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1901,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.32851687640321237,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2032,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.3408106398041139,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.17,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.3446202759269793,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.197,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.31050207208691405,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1753,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.39378625230060826,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2018,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.30665329808211034,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1656,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3270261957467447,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.1791,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3431363575716594,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2113,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.29742639531492554,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1776,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 1.7338351305279753,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1826,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.34108899809827836,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1753,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.3315035259520794,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1905,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.34017193714519683,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1889,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3155078665739788,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1843,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.348324574672238,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1887,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.340005247445411,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1983,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.311849244957659,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1894,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.3142890707760024,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1821,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.3204988354079931,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1935,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.28928003267192237,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1815,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.3243168008690955,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1775,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.3135826097630704,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1904,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.2892132593488121,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1673,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.3057162025269046,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1845,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.31959571482048965,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1906,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.30401396121639074,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1992,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3193667846199884,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1796,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.5581639590383644,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1885,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.3098210919416149,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1848,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.8661746102441483,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2051,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.34519565427090904,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1777,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.3531709360965766,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1732,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.3073283765058275,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.195,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.33680252034911146,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2005,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.30186701813668737,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.1761,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.3256311097424342,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.1967,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.28375531822027045,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1674,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.29057986486152887,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2027,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.3329187191426625,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2116,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.2992730538672029,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1764,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.30851399257608236,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1969,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.29882180950402126,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1637,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.33333716967902893,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.1747,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.2940760760278026,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1945,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.2945659643219517,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1878,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.3225604809928214,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2007,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.27861976735669153,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1761,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.32269146650626945,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1965,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.3205809178062454,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1903,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.31460559457213916,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1998,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.31191096686179826,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1824,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.2801696785472397,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1875,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.2871949564500215,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1781,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.3210477505110965,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2018,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.27485082112737386,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1689,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.3397790720890716,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.196,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.30834335656688516,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1991,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.2861781497982312,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1658,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.3446323849335004,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1958,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.30246619347222053,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1803,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.31950515980912386,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2005,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.27456117420459397,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1644,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.2988731030924335,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1983,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.3162508687245915,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2051,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.2824116481029432,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1854,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.2908736065759825,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1779,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.3016323777332082,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1825,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.28262233202402653,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1733,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.30728302413647596,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1913,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.2887593280672079,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1776,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.304290487932744,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1832,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.30203552059055067,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2058,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.27082533295808214,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.162,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.2939565008935577,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1937,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3000623559374984,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1926,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.28408713923078865,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1885,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.30448651020850154,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1889,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.29623579002073347,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1913,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.2586862497571329,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1749,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.2667774734609357,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1809,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.2942714244105404,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1994,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.3137174436892127,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2035,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.28033530557265623,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1692,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.28072946339809235,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1869,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.3188582979726197,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.204,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.29102541438022417,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1819,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 9.08310809640678,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.3898,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.287003575348811,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1665,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.309745871440303,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.202,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.2803453677808365,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1869,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 3.600848750333951,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1948,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.32749168993163236,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1828,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.30129693805641894,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.187,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.28267285850232093,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1987,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.29459939888198095,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1743,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.31193326713723213,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1882,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.288171515430097,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1856,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.3099645811489495,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2007,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.3005424016610262,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.186,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.2700854690796076,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1815,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.28087297982914905,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1818,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.2898866694920819,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1829,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.312567651410041,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1997,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.28349773483317614,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1917,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.28269083928253025,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.175,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.29731907116565826,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1717,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.3193554987854979,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1954,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.3216735956647062,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.19,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.2942322166059868,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.188,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.2934636186760965,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1894,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.2840504783503479,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1854,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.2796613578339925,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1775,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.2617752189664781,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1726,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.274845910326918,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1819,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.2953665505829677,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2008,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.26193500807432046,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1709,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.31926623486415856,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2227,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.28357308363223194,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1844,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.28416995051041194,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1957,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.28057085789136166,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.1753,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.30122720949736065,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.27972923691418317,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2001,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.2902554084702762,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1944,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.3088244074628815,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1933,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.27758022102875074,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2026,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.2656881387372324,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.1695,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.2811778996428667,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1856,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.27832364948465793,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1801,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.2964840206005331,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2008,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.27139541554563273,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1925,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.28788649115491016,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1921,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.28569099672202614,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1844,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.27301718211255777,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.166,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.29175601969759773,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2017,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.2707887595501017,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.166,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.29767962348758714,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2043,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.2603540209415978,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1666,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.2765977010384274,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1991,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.28022845310781686,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1815,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.2617219035783885,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1714,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.2967651339168811,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1848,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.2786089001586518,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1916,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.26353625263936054,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1749,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.2854472968639981,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1925,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.28366059201717697,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2022,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.2713306307898273,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1901,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.27573945377072356,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1846,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.27328825999021616,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1936,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.27834973168348487,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.18,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.2954578532531642,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.197,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.25980388349701344,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1677,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.2916890115962819,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2035,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.2917219520324014,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1939,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.2706549964292293,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1724,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.2914822597214419,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1977,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.26467194085288276,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1741,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.2777463015328604,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1942,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.3131116366998748,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2008,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.29258501122686636,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1748,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.3031626838787926,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1844,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.27156355712291974,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1619,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.2964270423223156,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1931,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.2926704224149585,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1923,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.27338203551305246,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1709,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.28249205654651244,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.185,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.28225543505950906,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1897,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.27589781455916684,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1827,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.28690445416232746,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.189,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.2921811253260941,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1838,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.27692110681704735,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1814,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.27114005050014944,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1903,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.2662430112958849,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1784,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.2908749916748022,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1892,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.26278829515281116,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1693,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.2751923014448542,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.181,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.2782211852153529,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1795,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.30058117825509284,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1841,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.2986070751354102,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1921,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.28199092225053435,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1924,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.2653649563981435,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1807,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.26693099180814017,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1766,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.28745369710860635,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2078,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.24409263639643428,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1614,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.2754478033804076,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1853,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.2659559278489075,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1885,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.26633796066871956,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1908,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.2631451433964006,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1786,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.2705037784370899,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1797,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.2978716086426158,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2037,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.25666803691404494,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1825,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.26889450608007937,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1848,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.2694765253291577,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1801,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.28233676226439486,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2122,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.34636783488310213,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1752,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.39016639577294854,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1344,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.32761218487761296,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1086,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.3063314472302849,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1465,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.2657486485647677,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1266,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.33836772304583324,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1265,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.3975142917770867,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1307,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.39667044884645175,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1197,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.3469390604458904,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1244,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.3020915653136125,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1337,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.2924498218188763,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1324,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.31038867555338434,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1376,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.3188256706931733,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1276,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.27832015364641505,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1147,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.2914475978133442,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1309,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.268230045192945,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1262,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.2793525195532585,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1305,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.2825775862355932,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1352,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.26370721110963524,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.119,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.32058696765765077,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1435,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.2679575651708517,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1189,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.2836043985048485,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1383,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2651317518184927,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1196,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2564639777514997,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1215,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2760482178921305,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1294,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.2736737858449901,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.125,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.27226681790235774,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.135,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.269300934178166,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1253,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.2741122405010604,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1283,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.24274288353781748,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1139,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.2681182256123144,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.136,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2701829519516294,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1238,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.2723133813159975,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1238,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.2515896088779529,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1198,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.26456376623025984,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1301,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.2588522514516872,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1281,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.2572162905894036,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1277,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.2612353306371383,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1285,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.2556536938295609,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1225,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.2573002938796664,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1158,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2772656758059856,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1312,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.250337585878379,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1142,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.2573529595584107,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1209,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.2578892549876232,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1321,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.25164893132456273,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1263,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.26115937315476107,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1332,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.2814592989796876,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.145,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.2414983686369148,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1182,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2533114835447423,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1221,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.2425567027804256,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1212,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.27456752153462965,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1435,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.2424361197208132,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1063,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.25983258480618715,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1254,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 11.475991267364762,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.29,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.24009598790682513,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1242,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.26369051910643876,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1398,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.27232532066689086,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1384,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.24014520344242804,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1194,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.2517561020982486,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1224,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.25149212058632747,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1288,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.25227660364666854,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1275,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.2412030002015276,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1211,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.2519533396853747,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1216,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.26679402917289,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1335,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.24629401628318415,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1176,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.24885832512804398,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1257,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.2633960793320241,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1218,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.27875616260932823,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.135,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.23954174603712045,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1224,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.2508934226053851,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1221,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.2603761125174849,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.126,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.2573293457257286,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1277,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.2590275656930423,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1296,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.2525492312383627,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1326,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.2627214882446801,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.128,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.2507218426704263,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1248,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.23601873442221094,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.118,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.26415360950346334,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.136,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.2452456275972099,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1131,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 12.844406164780844,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.5969,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.25481488724916335,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.123,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.27299640225546307,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.133,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.2549036632912028,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1312,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.23220650882326985,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1134,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.25881739045725244,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1316,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.25115281364086334,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1221,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.23761410683629589,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1078,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.2549908757365328,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1152,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.25553974346958996,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1254,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.2640432205629383,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1352,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.26459713544588775,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1395,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.2546337002894101,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1202,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.24806504917463204,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.116,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.26066152212773175,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1301,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.26613134346666467,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.126,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.2759089943001717,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1281,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.2574202054567857,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1236,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.24326171709085248,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1156,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.2651787227903902,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1279,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.2523703771927181,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1183,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.24661326722616309,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1169,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.25323527961353187,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1268,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.24406967265776858,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1166,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.263436691791212,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.137,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.26314638857191036,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1328,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.24609120630694065,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1199,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.24620818556537685,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1163,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.26459481374329785,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1289,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.25857783598767403,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.12,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.2721158363424404,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1315,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.24985657755181429,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1252,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.24201627286491054,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1183,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.25139281106758804,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1255,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.25767873413811737,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.119,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.2610288501056439,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1277,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.2473135713745857,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1216,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.2676999806665923,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1437,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.24566034319170663,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1115,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.27571182579885567,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1231,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.2463081608032069,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1156,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.2591762986908007,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1331,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.2559867182473054,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1229,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.28714043533915523,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1278,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.24966414497673034,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1196,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.23721458303391876,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1121,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.2578022307655587,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1357,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.2295878438790217,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1091,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.26176370085244194,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1323,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.24455675953944073,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1159,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.26846444928604846,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1374,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.24802312007755123,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1249,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.25156524184407647,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1287,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.2637114433062953,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.133,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.27594305816940423,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1381,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.24603059415656311,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1219,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.2580108286764703,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1281,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.2595245835995902,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1344,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.26207590341924897,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1396,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.24238346959136461,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.109,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2881925026844223,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1467,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.24945532891236807,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1208,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.2533687068651143,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1265,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.24002281103563713,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1121,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.25890529799729856,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1327,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.2363671587661075,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1199,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.26510994900881085,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1426,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.23312764725915033,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1193,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.2557357406640208,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1378,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.24996931978209633,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1262,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.2389090930943906,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1234,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.24557179937952256,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1247,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.24958582411468644,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1221,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.25210146511611337,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1194,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.2598548171448105,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1311,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.24500087845482021,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1202,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.2335893407762203,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1234,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.24420381875951155,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1287,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.23992652092450703,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1202,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.25146377311366974,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1301,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.2571912036160111,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.139,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.2499147631626967,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1309,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.25447511514318044,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1265,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.2387279508132382,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1177,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.26926704747919955,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1486,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.23234099399309668,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1173,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.2451207342225497,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1297,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.24373410728754166,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1225,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.2466292834489874,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1147,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.26869685062379933,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1396,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.24693097894827631,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.129,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.234247399628424,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1174,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 2.091407169650801,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1259,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.2577445919531224,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.135,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.2597821264908777,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1244,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 1.9140056069572968,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1484,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.23420324955985547,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1066,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.24308107529820086,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1174,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.2694171782085937,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1461,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.2765923441477599,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1325,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.25242793069143094,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1133,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.2698669522889779,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1113,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.2497410687388911,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1178,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.2512864789275847,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1278,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.2407592229265814,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1128,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.258641495447087,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1272,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.24460463802589472,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.118,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.27006608157826945,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1396,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2546126898420998,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.126,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.23529542877064252,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1129,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.2461075920231223,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1268,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.24473310264873083,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1281,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.25417290277180304,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1296,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.24917897541081496,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1271,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.2540001859369339,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1415,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.2498972947978466,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1307,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.24756152200306325,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1199,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.2622067060549322,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1358,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.26592964076933456,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1394,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.24147537870372252,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1202,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.26885482124097754,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1465,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.24915419602887995,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1267,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.24942540775568325,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.122,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.26724467552628567,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1401,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.24279178783690425,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1173,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2540727762494961,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1218,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.241727708409229,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1207,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.2569289181832268,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1276,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.25030780257424684,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1398,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.25422603983369113,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1346,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.24778307736576013,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1233,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.2540025279692511,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1342,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.23317797644746624,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1175,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2286684082639265,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1115,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.2656789041393987,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1306,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.25033318821611944,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.127,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.2466712553622524,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1338,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.23379999783327002,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1136,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.23459771710762586,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1176,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.2679889375573025,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.137,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.24400047999353044,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1173,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.2500022607216011,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1202,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.2394595188037875,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1127,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.269335971840042,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1373,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.257075155102937,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1205,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.2615377762933465,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1305,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.24107623041867188,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1348,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.23736887780517818,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1141,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.2469491590842182,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.124,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.2526422470572365,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1351,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.25947170165608663,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.128,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.23840610047324956,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1169,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.2561248700445423,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1296,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.25248386616856167,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1329,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.2394506108927754,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1197,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.26145628781702135,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1238,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.24726549039586287,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1172,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.25172011585511367,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1187,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.2407352062374037,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1145,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.9255699154956983,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1473,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.25033805801068787,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1213,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.22851681754778708,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1197,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.26418835146372827,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1366,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.248830961016473,
+      "learning_rate": 0.0,
+      "loss": 0.1251,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.4155535887774515e+18,
+      "train_loss": 0.28850433091080285,
+      "train_runtime": 124436.168,
+      "train_samples_per_second": 1.27,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "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": 2.4155535887774515e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}