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": 8.779956035991004,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.2068,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.876345752758109,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2128,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.555597493154622,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.194,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.376198663931712,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.203,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.170695850346577,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.1931,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 8.043555009041624,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.2043,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.805790828585471,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.115,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.067337946103351,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1018,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.1665083089852035,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.1022,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.9782774374004166,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.0465,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.969239630984209,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.0827,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.9977936102248304,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0242,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.097507685858799,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0062,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.5629682860881915,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.0062,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 2.9004667222500493,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9765,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.4696769230767144,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.9539,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.9622320838749734,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9456,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.6001467209109475,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9173,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.8665739322919257,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8961,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.2667071559082108,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8805,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.4273544578175439,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8827,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.613118678296495,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8439,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.5220975963216765,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8709,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.2770697721331914,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8689,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.269576785442013,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.8565,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.0984675819434921,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8491,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9574052604640764,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8226,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0934071365323197,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8196,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.0104480776723452,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8171,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8757623609269412,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8054,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9189656743439127,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7914,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9003688105053889,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8103,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.9679243274012304,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7982,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.9342804542520795,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.786,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.753985042492855,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.782,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 1.0344376496603689,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7987,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8199194774393387,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.8006,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.8132240684401271,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7842,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.9590269486737395,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7891,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.8635268056895001,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.781,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.8098266198001376,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7709,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.7535540421213317,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.7835,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.8902535142919733,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.7898,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.8608261132869378,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7547,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.9587095102518554,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.779,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.9395058177559364,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.764,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7980700982763563,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.7531,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 1.007801530371745,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7849,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.9296620407687815,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7688,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.0660355164719828,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.7772,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.927812048858375,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7657,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 1.0810978476608,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7573,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.9180295843914953,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.7383,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.817662400495055,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.7666,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 1.089649063146894,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.7556,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.9365441197277795,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7559,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.8760144188721736,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.7561,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 1.1938808047481828,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7523,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.8243800086377359,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.7536,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.1080426038313245,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.7826,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7118806401181211,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.7534,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.0212059157745246,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7413,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8710799761849832,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7558,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 1.1930294947719235,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.7625,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.9963286158989368,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.7458,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9438853862438951,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7356,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.2216921631386153,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7493,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.0748352956036906,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.7587,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8615743290367139,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7452,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.1157312421787036,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.7504,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.5679514844004019,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.7359,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8792866642426611,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.7634,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 2.296416933640252,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.7403,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.64114792653819,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7496,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 2.1068890652624104,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7549,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 2.192045782226463,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.7548,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.3445180562716705,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7444,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.7396439789138016,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.7254,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5564203877282117,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.7567,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.5094943853785008,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7437,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.5512494461727726,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.7411,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.5310188962825326,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7634,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.227059269612612,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.7226,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.3734646788997478,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7349,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.4858252983796696,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7412,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.2975317794862558,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7337,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.490062041439045,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.7345,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.2826093120766375,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.729,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.23461069929458,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7255,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.1921931256906344,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.7437,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.226380198922123,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7311,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.2112937013538678,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.7507,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.0474857378030198,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7485,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.0054786846469825,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7303,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.2358677771183888,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7452,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.3580973089303647,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.759,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.1934983999779796,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7445,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.4987949076754632,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.7495,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.186528901268842,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.74,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.0469792462922745,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7375,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.5589162743094263,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.7501,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.9846242834222967,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7426,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.6506608886453817,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7371,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.18528027181501,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7441,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.3836331123436327,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.7168,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.4839970407336533,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7429,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.0720109596095537,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7278,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.406730158628717,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7402,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.07121758690747,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7428,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.6565254101503806,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7388,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.2814516453611031,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7273,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.4571472008979245,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7558,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.6138090768523936,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7507,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.2230069997294344,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.7345,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.44803200226667,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7409,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.4697703174965504,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.7402,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.1820707900916225,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7425,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.4022327767388012,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.733,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.9686776923795197,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7297,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.3551119842736807,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.7387,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.1356673445543406,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.7264,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.3196950151886542,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7087,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.3939468285988323,
+      "learning_rate": 4e-05,
+      "loss": 0.7547,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.9117395943007357,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.7214,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.5261420257499574,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7355,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.1803308475823788,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7292,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.5820312253848083,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7394,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.7161760574285732,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.7504,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.0139760986571027,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7486,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.5022833630679242,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7299,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.5838794048976121,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.7533,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1639519850732971,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7395,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.1037606247790588,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.7379,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.2363618581110816,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7185,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.5564016642625973,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.7419,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.2477265995191438,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7436,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.624933626190832,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.7394,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.718839193682717,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.7313,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.5925320219676011,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.7258,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.259335905299808,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7523,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.111608631212592,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.7294,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.8733784722751534,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.7305,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.2349851543391839,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7321,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.5525557836611434,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.7279,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.1937630998328554,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7231,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 2.0646269513740907,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.7355,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.3466983046200698,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7275,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.8976378663458129,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7322,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.4088331105449101,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7222,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 2.0228436623637025,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7491,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.398302750747318,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.7288,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 2.3452139092456172,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.7447,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.9950887418958936,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7158,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.7387426651220086,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7173,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.7486786760637711,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7224,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.3784963708879512,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7323,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.272250526702518,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7185,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3293616742081507,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.736,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.1481929030426052,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7299,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.139413434428375,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.7283,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.1586052306371992,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.738,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.0324214656040507,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.7319,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.0341859320025595,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.7254,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.3316407825668917,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.7338,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.2352162468427053,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.738,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.1373026272625086,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.7264,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.1361094108710101,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.7176,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.2564014645648387,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.7231,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.941265422291968,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7356,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.5322208558444614,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7295,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.9975528834393993,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.7255,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.5611745696327786,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.722,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.0040839214751132,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7266,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 2.036040535007794,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7202,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.7247632912731627,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.7509,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 2.125441278882217,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7523,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 2.277371964339174,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7511,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.242775258015007,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7454,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.8392995709595936,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7347,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.4512044318923176,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.7243,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.6787297332942794,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.734,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.3085997527237643,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7428,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.3570517549883514,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7126,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.2167672647130794,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7345,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.4130504143971376,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.726,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.4508572622344997,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7194,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.4578174766905472,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.7162,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 1.4824176509179976,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7065,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.932431192012315,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7229,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.4280528104803247,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7293,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.0117123548727867,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7181,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.3632208172096891,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.7199,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.1167704700277867,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7154,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 1.380816828590691,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.7181,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.1797633059034034,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7188,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 1.5458121682929669,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7042,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.2140491033197818,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7366,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 1.8390204171927953,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.721,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.2566700996382274,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7454,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 2.0252202176284286,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7237,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 2.1800781839961902,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7156,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.988416451703304,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.7091,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 2.175309357646625,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.735,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 2.0015038213167613,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7227,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 1.3573703891157403,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7343,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.40193392594398,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7151,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 1.4164967845493006,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7273,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.9221290874321839,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7257,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 1.244959191432455,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.7276,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.368677333611406,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7357,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 1.0174896179403994,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7298,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.2981600891330856,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.7332,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 1.0309410492085818,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.7296,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 1.0794664520397232,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.7045,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.3572057293793132,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7259,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 1.0808683766288134,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7196,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 1.2306178769649931,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7262,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.0221936472110447,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7248,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 1.2599200840445346,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7121,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.8670033262990499,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7319,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 1.1436210479935918,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.7187,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.9206865639083077,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7203,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.9594318002116811,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7188,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.817150026173472,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7031,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 1.202322435198687,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7269,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.9642905023390675,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7238,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.9156750899925488,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7264,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9495610898081978,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7007,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.9201397785172517,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.7249,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.7807853659312467,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.7181,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.9678711341267278,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6937,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.8791465256256581,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.7396,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.8795982463023685,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7194,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.8482647804904949,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7024,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 1.0752955163584783,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7156,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.9804980383433678,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7085,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2497896191158633,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7228,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.9765568068169652,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7198,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 1.2140759443165556,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7185,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.9067999243717847,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7039,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 1.0060492345514298,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7161,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.0538804610335093,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7219,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.8202332940965192,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7167,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.1762620420745302,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7086,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.8288356056801123,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7099,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.8964880787622901,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7277,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.9771876772608222,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.8015,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.9220600187049536,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.577,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.8091234073660893,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.6363,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.0521477334201688,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5959,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.0187238203925986,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5975,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.9612767480836649,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6009,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.9839427589171643,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5574,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 1.0605377048567441,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.6709,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.9218774458628036,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5828,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 1.068977101436574,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.594,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.743483352485253,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5886,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.9024776526166471,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.621,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.9208762184747297,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6097,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 1.0377567501219147,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5916,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.8752768918750023,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5862,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 1.2487646375958927,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6119,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.9667801915555777,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.6115,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 1.2122306797063496,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5666,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 2.481787404719381,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.6177,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 1.3809874728029679,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.6055,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.8445180940375312,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6212,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.9146043399944691,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6431,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 1.199521045451661,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5681,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.8686447791533164,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5942,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 1.320488882471905,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.574,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.8289468719000648,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.6189,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 1.17991382847411,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5895,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7735359339473463,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5865,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 1.0597397862240852,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.57,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.8303103612606169,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6204,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.9948456031120342,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.6017,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.7830666548187155,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.6084,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.9122743487034625,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5937,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.8341165041438009,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.6077,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.7938696723254358,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.6061,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.7977395625226257,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5733,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.8577582343661107,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5834,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.9173497925099113,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5818,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.8196011477603284,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5926,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.942412940302146,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.6136,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.917755573743344,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.5901,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.9790346493433523,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5917,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.9369071713352926,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5951,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 1.0889666224209016,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.6465,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.7669790350823473,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5648,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.9329100267912197,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5856,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.9162206917450072,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6184,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.7301380239554629,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.6039,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.8668010114409725,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5895,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.816900779631978,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.6192,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.7783323833461534,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.6005,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.7817799673369924,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5716,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.847415686765197,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.592,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.7901040701314918,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6063,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.7913943122103162,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.6023,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8423449092605542,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.581,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.803495872168933,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.6336,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.7983509173668089,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.633,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.9050272986640039,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.6277,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.8362959506579118,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.6103,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.9497578268370876,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5846,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.774366232068752,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5737,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.7869245241041276,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5885,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.9519637377625481,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.612,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.7607800344654675,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6124,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.8338972491830801,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5706,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.7256794545554963,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.6149,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.829241455069063,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5999,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.8335269193795395,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5961,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7088734701573679,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6078,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7439505678047068,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6006,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.7733910646209663,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.592,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.825816113810982,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.6146,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.7548573263642249,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.6121,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.8380138587669078,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6163,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.738832579840157,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5836,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.7050532623636397,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.6192,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.8191874343371068,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.605,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.7522921106122592,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5818,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.7746986772154777,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5922,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6993005112132626,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5986,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.7422480761266254,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5818,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 1.217309944101288,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5959,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.8433098821581096,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6317,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.7841895328974249,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.6138,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.7381384724824026,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.6329,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.6896044768187212,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5731,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.6686903174517975,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.6035,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.7527691827406118,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6135,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.7316565426485271,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.6145,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.7605941071115156,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.6253,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 1.0672104537495104,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.6179,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.7147555436558348,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5731,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.9116768029590926,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5999,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.8618033807748221,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.595,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.9850197568497383,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.6131,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.7693390280153322,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5958,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.9502837173553585,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6117,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.7732744040155568,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.6099,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.8721318896430886,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.6335,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.8351819329362281,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.583,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 1.1093838693303426,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5882,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.8380259574425097,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5984,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 1.135405215850108,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6276,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.8643416579902865,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.595,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.8744381243994946,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5605,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.8965113266586198,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6173,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.8658043660742726,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5828,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.9003303422716746,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.611,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.9356037282747536,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.6087,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.6937845186822923,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6068,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.7674023176915108,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5801,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.6671904746607676,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5833,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.753518678539948,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.6113,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.80684468138824,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.6249,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.6885405561987277,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5986,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.8442308124047975,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6007,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6683467010688418,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5658,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.8593209859531321,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.6373,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7837557688414466,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5766,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.8253698863581986,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6483,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.7192472993475514,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5833,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.833980805005142,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6296,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.6783107086802872,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.572,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.7917965853574943,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.6315,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6761410299131444,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.5526,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.8502222830451954,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.6489,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6922012792174727,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.599,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6569898088460497,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5892,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.9194062263380882,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6224,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.721124209966329,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.571,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.912211780849037,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6079,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.7400210154350522,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.6234,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.8861326251085625,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5893,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.7565884433731603,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.598,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 1.0615720756187228,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5758,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6415696097515978,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6021,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 1.123554116759346,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6427,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6949559825905122,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.598,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.8569381878861686,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.639,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6541552191925484,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5805,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.7210740304831861,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.6062,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.6672929859862741,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5642,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.7876186937328371,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6309,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6742606500738212,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5659,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.7981793040310431,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.586,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.7773997631911811,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.6222,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.8950041800548644,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.6073,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6698445070338004,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5791,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.9640730428265056,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6233,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6845548229228348,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6152,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.7504335185761651,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5615,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.8183585494256541,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.6213,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6283858859657471,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5958,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 1.0688514674800491,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.6051,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.7599787654128091,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6228,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.8323393590729463,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5839,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.8137412663224808,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5994,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6066347700054976,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6044,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.8004896219821963,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.6374,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.6948901119030524,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.6236,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.7801773283543758,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5826,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.7103851773225611,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6069,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6830428871016865,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6048,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.8225668252947116,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.6181,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.8160373701297735,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5923,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.6963604261305352,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.6207,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.8283824112500815,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5818,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.7634225830945287,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.6179,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.7581622104171019,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.6289,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.8723291812206244,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6063,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.844457290753851,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.6125,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.8071183905421511,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.617,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.7318698801222467,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5564,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.7713667989497932,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.6045,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.70031132104932,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5973,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.8538736380869427,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6023,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.7586182762672774,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5695,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.8094903637263637,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6102,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.9313770769126226,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.6157,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.7006682479037436,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.6053,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.7854099882372378,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6028,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.7763458467477186,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6536,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.800516175978354,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5549,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.6592699639142288,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5732,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.9450738706292406,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.6442,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.8135532205221871,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.574,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.8805624192182497,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.6212,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 1.063456475832844,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6265,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6418283973294452,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.5915,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.8945455394337996,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5941,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.7310537772660811,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.6283,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 1.0223303025156067,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6231,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.7720104357684615,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5609,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.9220911774681197,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5864,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.9091107506962446,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6403,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.5985556028687398,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5857,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.6959655985243266,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.569,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.70567401412607,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6067,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.8950346509923266,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6344,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6208279009221653,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5909,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.7832139494558927,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.6193,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6979654570894746,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.556,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.7227501592837257,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5764,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.8540539356115663,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.6099,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6426553985650663,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6127,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.753365786635519,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.561,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.8621142779499456,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.6818,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.9317956196110073,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.5892,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.7248587567107645,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5563,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.8627254670842393,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.6137,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.662619413296214,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5837,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.863264246311926,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.6018,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.8503425907698853,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5914,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.6760853443496898,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5664,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.7987484183142147,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.603,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.9348151334962516,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6216,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.8198580671411968,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5621,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.9535281598387955,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.6452,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.8680787659725386,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5871,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.8058106079760238,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5913,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.7287871871100913,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6158,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.7623519207130958,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5986,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6883141842442756,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5936,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.6029709941244933,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5818,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.7064394247810619,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.6015,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.658207977183432,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.6288,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6625897539609996,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5701,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6855186979842101,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.6055,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.6113469652262091,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.6288,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6600105456242685,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.57,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6586786361016082,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.595,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6252707381463793,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5931,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.6774168340824718,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6489,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.6263458653606698,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5591,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6990647495492194,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6324,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.6384052938584556,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.6301,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.6399079192824985,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5618,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6689771443768475,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.6056,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.5898876195222422,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5931,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.6766970530286623,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.61,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6058989252693917,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5845,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6592896162657567,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6077,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.6325777504013063,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.619,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.6155849213451515,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6042,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.7312285564813927,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6278,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6181899884433903,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5848,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.8060527612386164,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.6029,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.8076411150867309,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4294,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.7235472959577052,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4357,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.7913118170074814,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4053,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.7041279028943295,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4158,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.8305554362537264,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.4218,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.7665897246679679,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.424,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.7247098657478236,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4266,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.7849664140605209,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4344,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7222056117847149,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4263,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.7223092565193392,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.387,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.6885442370009542,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4137,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.6793330751653218,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3961,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.6590111724903747,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4441,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.7719252297592335,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4466,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.6264241709370291,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3895,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.8184453582693845,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4489,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6045123652876392,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4107,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.7922611966277644,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4255,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.6189274864608124,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4228,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.7863898433507917,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4046,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.619179088814123,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3679,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.804322373974654,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.4154,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.631908059462753,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4091,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.7127152377504042,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4181,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.720222006494778,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4393,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.6190434156436312,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4001,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.8354104083214744,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4709,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.559148367147194,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.396,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.6295988078030835,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3929,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.7556461999940678,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.4287,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.7026270879719,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4306,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.697367325706688,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4048,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.8121632302200471,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4543,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5622182553775033,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3671,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.6614227171966519,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4339,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.6632832099358906,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4227,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.6158229567231566,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3846,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.7461838450549578,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4674,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.6559610076552819,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4173,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.6938164430853851,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4253,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5466129700447683,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.431,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.7611307293260523,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.4019,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.6524587437926467,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.441,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.6202892668675588,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4005,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.7872243446700788,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4302,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.6397624498888917,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4135,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.779008777781057,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4373,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.5514628448684552,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3868,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.7159112574262776,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4357,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.6564180144196786,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.4448,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.5477622401504767,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3936,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.618223565352187,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.4137,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.6264005025550942,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4431,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5797755189187694,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3832,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.6622580551466517,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4266,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5858357165720357,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4059,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.7237060373117518,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4432,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.5436836842285193,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3945,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.636079216022333,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4115,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5727111226800747,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4465,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.7297491770316885,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4164,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.6114832405965434,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4066,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.6803573621228356,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4661,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.574611883552285,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.403,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.58657692193652,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4126,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.6379428283570656,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.417,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5675076576131323,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4006,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.6235507665611633,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4687,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.6380740578065469,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.398,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.6024140044458501,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4423,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.6370732901323515,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4335,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5617068470831743,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.409,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.6172716384755423,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4241,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.567709778434602,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3909,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.6172568468448479,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4182,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.6557846446262274,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4135,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.5926822034638828,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4421,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.6813730993108139,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4125,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.637973646511557,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4472,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5743063663050034,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3945,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.6268190990358128,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4581,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5884839399461551,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4182,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.6892254861753139,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4562,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.5860041702301577,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3922,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.5481349244614959,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4038,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.69552677038474,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.42,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.5413541380938526,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.4497,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5632870194169036,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4031,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.6665168876499463,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4385,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.608255470679372,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.4172,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.6302085946147198,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4556,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.57288592398841,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4095,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.7063283742599255,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4571,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.5551469947700216,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4033,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.6102512606127177,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4482,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5850937674378319,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4266,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.6959807192772832,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4333,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.7013879039620231,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4338,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5659161182606123,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3873,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.64926696170659,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4388,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5750944643238565,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4447,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5946234695216703,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4092,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5947837488529165,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4237,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5554496876832564,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4199,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5982468325684853,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4247,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.534012315313601,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4108,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.6137458248020825,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4252,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5765365057065787,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4427,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.6762969744940368,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4071,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.498186064583834,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3985,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.6218814237418451,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4674,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.569721840308275,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4249,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5503438161156198,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3794,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.6003501169402847,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4442,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5474006111607066,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4248,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5749286977256908,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4209,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.5943834263944185,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4692,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5463265025840179,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3947,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5180820040122553,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.406,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5761673479137573,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4615,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5506136448136051,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3997,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5900584955596591,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4131,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5849150252746447,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4053,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5428608189124354,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.446,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.7309967308385624,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4461,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5916952451806154,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4761,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.6802726541676968,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3863,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.6011325988758737,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4234,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.7563496416881336,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4751,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5013098647277986,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3874,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.585409545087738,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4266,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.6333877405069618,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4223,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.620161452754152,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4296,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5498685829072283,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4047,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.6150306097119834,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4421,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5207785071040179,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4136,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.519436745351888,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3858,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.615040686095088,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4549,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5462404679260995,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4327,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.4956731075781798,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3935,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5713922475023416,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4304,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.5648944159147253,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4246,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.6145174515477833,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4579,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5339543856917675,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4066,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5810072434534694,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.433,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.5609512142388785,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4034,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.6154355778451599,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4124,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5463530327204988,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4066,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.6477352484113624,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4492,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5551482802916599,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4264,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.6078241196407503,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4198,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.6147025994111255,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4324,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.563614834268323,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3947,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.7343561010312619,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4385,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.637203172028446,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4243,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.6007873535567125,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4468,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.6709315312229678,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4153,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.540905543325581,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4112,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5823437322526874,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4205,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.5772418280025476,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4102,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5729990875365113,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4447,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.589731120384547,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4029,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.6621972728158764,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4518,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5416473235679246,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4194,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.7416135919439327,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.415,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5789693345984614,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4505,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5981898082092604,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3917,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.6751932846038735,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4419,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5740910687179637,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4148,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5271555826517297,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3893,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.49532167866941423,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4172,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.6363732387220368,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4227,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.5639214991171093,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4316,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5212314703496558,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4048,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.6223844386330251,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4628,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5289698137166597,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3897,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.5395853311806736,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4749,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.5130069792560762,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3951,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5486889029191206,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4339,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5224738250273121,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4172,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5226945031012461,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.402,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5916512446374997,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.433,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5664479267555284,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.461,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5030086526456734,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.39,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.6472745550586063,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4431,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5458525007989735,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.425,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.5338117182341229,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.402,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5759999550939722,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4195,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5846606422560572,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4215,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.6062273334930762,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4455,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5629164219454885,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4009,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.5805081462374111,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4181,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5516244959091103,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4542,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.521330391823812,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3911,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.6414458496758074,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4609,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.5412706415994744,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4025,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5955007206105106,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.4129,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.6033375947920819,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4157,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5513708831128076,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4223,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.5978441122860034,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4438,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.6277946251024782,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4521,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.534692738257971,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3998,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.5264430928723646,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4425,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.5739809679878654,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4482,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.5196523085749417,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.399,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5401004536942666,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4219,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.49728768939905565,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4095,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5242749806072996,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4253,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5523019939598484,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4323,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4969087390916471,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3996,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5364393743522541,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4142,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.5401811983686314,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4301,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5259568933492539,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.403,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.49967860103160716,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3986,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5977172291061399,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4747,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5736572962062734,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.4258,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5260738185274342,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4104,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5342591959883249,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.404,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5489587705053203,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4141,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5690341015657058,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4167,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.6221721033233532,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4397,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5217207786163669,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4532,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5091868484603379,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4011,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.5602583317331565,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4134,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5204088602865394,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4159,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.547056563273602,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4236,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.6349661375720385,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4288,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5479404909170428,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3915,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.5509720142089507,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.4243,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.48568421141601864,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3846,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5327896401538186,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4221,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5313144122526576,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4522,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5008542184229119,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4073,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.540741661613399,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4271,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.5009889861920864,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3901,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.5724508457924501,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.4406,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5154842623015452,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3849,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.5222290548168318,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4141,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.49568805377718844,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4118,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.6020720895124108,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4329,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.5425331849810567,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4364,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.5016149313585949,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3975,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.5106862898923678,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4291,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.5212644570166306,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4169,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.5147272478435754,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4222,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.5528231215330539,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4232,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.653675623971194,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.4206,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.6767849526273417,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2499,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.5310350549594045,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2494,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.9029663643212398,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2704,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.8119471062830821,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2764,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.6524728048461756,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2752,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.6779458276271476,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2768,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.6133995085166996,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2281,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.6049615434822381,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2639,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.5556088332632677,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2617,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.5661467119151554,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2621,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.6180371146907723,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2639,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.6067514800503873,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2539,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5720644331799105,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2458,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.5155571996381435,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2429,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.6487414533630892,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2771,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.546497329917398,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2305,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.5744550440348769,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2515,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.5293778464255919,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2383,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.5872912555015652,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2718,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.5044565683727978,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2314,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.593702501998762,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2685,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.4837632042011521,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2343,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.48217774893801646,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2495,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.5173408034407532,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2787,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.5266211707165595,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2602,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.4732194695668499,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2527,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.5107643123453625,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.282,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.4889913558619808,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2536,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.5137940942151847,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2578,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4466711507071297,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2158,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.5373099347561252,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2957,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.46744264507490063,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.229,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.5149639473552924,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2797,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.49041590128161555,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2559,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.5324646234466369,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2395,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.4909361762444747,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2667,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.5317188497586223,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2559,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.5047780800082838,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2501,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.4888027194212344,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2414,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.49987180599852943,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2702,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.4934208182810149,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2433,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.5089830189912631,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2578,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4828795189445196,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2511,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.47371218161878814,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2412,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.5047238265689786,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2734,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.48146835952251915,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2388,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.51137895717355,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2614,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.5095889085010091,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.247,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.505772311521782,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2631,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.47958283545671276,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2547,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.5126373573999679,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2586,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.4733026115965674,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2311,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.4855556072296886,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2514,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4879660666794901,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2648,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.4881310614257284,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2888,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.48078651254347676,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2611,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.5036631886567187,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.262,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.5050855315653818,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.22,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.48726353322720195,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2619,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.45558688031332695,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.243,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.4889776611908116,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2776,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.47689711244662136,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2436,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.4539983434970746,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2388,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.5004946758178199,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2544,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.4817340590925396,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2631,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.4790734928097447,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.275,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.4580940025835736,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.242,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.49316782884648486,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2505,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.461636138893164,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2617,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.46301110526496997,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2417,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.517340048612595,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2516,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.4772540828683719,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2475,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.4936258339720747,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2498,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.4751003782062949,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2434,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.4739423828672922,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.24,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.4797060219913379,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2524,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.5069453708696401,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2694,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.4704780978869517,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2334,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.4751492339221502,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2534,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.5099258605728971,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2429,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.4810356448034204,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2349,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.5054675953069688,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2673,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.5007468524283758,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2481,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.44625319973562555,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2417,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.55374516180382,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3074,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.45215823125131704,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2432,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.4640837019718743,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.257,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.49210218348174006,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2471,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.5136261204760401,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2924,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.4328769333863554,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.2361,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4488492572935635,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2238,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.5133907779403278,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2511,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.5045558991891764,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.251,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.473710150060532,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2638,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.47784486242033153,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2524,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.4907215223012512,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2593,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.4479670993107804,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2527,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.46299740848441345,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2573,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.465273874316677,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2501,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.46813655852955083,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2606,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.4434909822702529,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2286,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.4792921447403918,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2453,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.4514903917254478,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2471,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.494709602748405,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2679,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.4811687636098439,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2507,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4562554098087837,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2165,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4665337302844658,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2392,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4631086992179611,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2705,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.46103094166756586,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2333,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.6583797932567371,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2723,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.4978389225463033,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2637,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.45203177153865454,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2404,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.46057137124980846,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2359,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.4762648136851233,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2759,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.4479882801781863,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2368,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.501217045492845,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2695,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.45323955874931543,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2377,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.45464231709187747,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2737,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.44443577720016686,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2266,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.47305888446762717,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2488,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.4327605473349722,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2343,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4543014145125534,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2586,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.5101431364847937,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2761,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.4238645358895808,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2228,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.4649482274347785,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2593,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.4530135186849534,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2602,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.4418933675673795,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2394,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.47468368884448436,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.266,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.47088206564179347,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2405,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.46114871078598996,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2438,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.4600978168330241,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.241,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.46510992095459236,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2649,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.45754421855554517,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2377,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.46255112001724613,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2591,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.44418641266360165,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2459,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.45125014498467986,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2369,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.45707704024525686,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2479,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.5054993220555059,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2815,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.44441705031382855,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2319,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4826536765438439,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2644,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4450029615689764,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2515,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.48464593991056537,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2668,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.4488667112276337,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2366,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.4317482800176409,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2317,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.5078405004786296,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2785,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.4772978896989401,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2443,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.4228138013865149,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2228,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.48537350996822215,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2559,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.4726677063766925,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.27,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.4469758168572851,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2404,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.4321010559870016,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2412,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.45222825611696027,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2617,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.4586506883833137,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2334,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4853952652642841,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.278,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.44039391859770166,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2511,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.4504453615987333,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2611,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.42666303952650675,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2286,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.46182489842342306,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2563,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.4440456816210939,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2635,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.46856640350320083,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2815,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.4751809740426853,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2744,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4479684557268931,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2472,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.4932735928343901,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2509,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.5411692675469305,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2284,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.45919812386622655,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2474,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.4410300767152215,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2618,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.4521502458544625,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2625,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.4290780291692042,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.24,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.47023448971841114,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2734,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.46225665506587005,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2493,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.4675223465320805,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2432,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.47062986799641504,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.2513,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.46165551802711097,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2533,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.4576553751580059,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2495,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.44048119145178694,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2417,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.448857029040218,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2426,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.47044828890467677,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2391,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.4727569093717149,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2591,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.46871346762134436,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2501,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.4810217652016634,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2502,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.4465398199444572,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2553,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.452310897195024,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2526,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.4551259095632874,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2676,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.4375254404587069,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2362,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.4667877439888901,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2531,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.44479543871865085,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2514,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4369989174356349,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2217,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.45470308044937296,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2759,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.4591041816977353,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2534,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.41627162534386364,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.236,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.4506015704493327,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2574,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.4466609336445167,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2447,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.5056024907110217,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.3077,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.40697816072519083,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2125,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4805648132528264,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2573,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.4300321785916346,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2142,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.441510223716156,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.219,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.476765292223157,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2686,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.4863525333470845,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2614,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.4330920215816956,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2246,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4493211067120216,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2471,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.5006709107032945,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2839,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.441007293998936,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2315,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.5102920217637892,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2853,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.46242178726957434,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2236,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.4605995147903185,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2516,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.43875163864986544,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2299,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.45924853366582225,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2393,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.45513690903143994,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2536,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.4785331441802964,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2538,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.46714562341070925,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2499,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.4533626630239424,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2443,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.44840421237480343,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2442,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.4724937599159171,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2651,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.5141948229401133,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2739,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.43362072055512324,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2452,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.45095899196319017,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2514,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.4463226053968478,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2644,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.45785687570982364,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2391,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.45515558035048126,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2539,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.448167645757449,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2366,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.44266428983672185,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2412,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.46842539521903837,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2726,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.4364471448913433,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2499,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.4513391178422731,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2472,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.4419806842174455,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.223,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.43880920577420707,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2329,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4367374978990697,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2496,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.4372152196172165,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2423,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.41074148582675596,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.225,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.4780142507920829,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2507,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.43585973902606756,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2285,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.45502135415604533,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2508,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.4566394405489211,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2531,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.480305317012437,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2375,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.8922754232820582,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2295,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.4637931201221718,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2461,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.42668158702084186,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2378,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.4423057429173872,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2557,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.42765878537090335,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2524,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.4399542718534751,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2364,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.4501640466380091,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2433,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.4476895845947317,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.241,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.46478415053826305,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2663,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.4506251771249016,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2468,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4848814311002056,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2666,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.48006982320943214,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.1907,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.6717508722883253,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1758,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.5743947114283303,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1534,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.4514873383773512,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1432,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.40630358772954106,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1575,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.4595830443803294,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1356,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.633057551466446,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1582,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.7205330571682389,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1564,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.5643688944760733,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1338,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.5637608928695861,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1666,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.45646202095570443,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1439,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.4603095110163774,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1576,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.4521416654333245,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1521,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.4772868495047966,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1524,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.5030209488518995,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1574,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.4593764513634037,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1563,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.4576935268810259,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1527,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.3890396641398617,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.125,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.4410787561010262,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.156,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.43135224236029784,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.144,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.48271120348900376,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1605,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.4429072599862144,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1382,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.47710000934922014,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1697,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.4066280687580227,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1277,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.4267570465581919,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1546,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.38767298352460394,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1333,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.41023442539743915,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1602,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.41486753914561797,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1407,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.3844903889195765,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1369,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.4326157692506466,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1665,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.4159421594293745,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1346,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.41349600470157183,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1555,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.39186702364180215,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1568,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.402462237599413,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1626,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.3696320131870171,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1383,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.42355438321695865,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1701,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.40566789866380637,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.148,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.3934295581205668,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1454,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.41247629354841137,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1605,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.3845001914703227,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1499,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.41334256788976115,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.169,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.378474895814812,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1425,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.42284544937035934,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1765,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.38537808338780194,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1359,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.4117315449563182,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1451,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.40077297450098825,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1449,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.3882396346175873,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1419,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.39301406412706574,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1454,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.39393117062092353,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1482,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.3874125847995182,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.135,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.3957189869760419,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1434,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.4024766896931346,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1417,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.3712590826211312,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1197,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.41641049337450436,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.166,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.38908829144670753,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1297,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.3973542993466034,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1596,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3674698500744939,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1287,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.4030749726428868,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1606,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.38864138947249555,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1476,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.39903760982356684,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1446,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.4130431161502388,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1477,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.3861810846492718,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1414,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.38278970758126196,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1433,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.377662234088879,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1429,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.3948009081824763,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1551,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.39472119880113066,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1427,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.3869216190560136,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1404,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.4075656549644338,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1555,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.3952905077860514,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.159,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.3769094684182557,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1503,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.37399807943948027,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1416,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.4003185209094802,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1405,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.46402764954762654,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1381,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.37566981210824935,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1387,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.3875936424066085,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1476,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.3805900698252471,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1445,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.37358186460052617,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1548,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.3764031970583431,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1383,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.38392878037570066,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1458,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.3844405366552859,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1399,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.38636561060855473,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1648,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.3595143430111566,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.146,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.44651627334108956,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1408,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.383820494519033,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1372,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.37172879439148804,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1277,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3923222029426957,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1621,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3736962609317007,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1339,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.42191547592742756,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1681,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.3884397071686507,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1549,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.39391748949953015,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1475,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.3755375823473751,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1526,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.3779197047444541,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1475,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.37114712167972835,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.147,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.38227364015331355,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1531,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.3592706655702296,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1336,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.3825453709716308,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1538,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3983980083687055,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1563,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.3754950298179594,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1404,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.36248173131086336,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1184,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.41994988104767084,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1671,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.3889658331768169,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1566,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.39533220432171273,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1481,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.3796775326104017,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1424,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.3928206957322147,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1516,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3551368262307791,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1472,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.40519715904794873,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1697,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.37303585076672563,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1402,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.36340410811283264,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1373,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.39994270936080506,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1467,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.3905422717033787,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1421,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.38618429869619175,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1451,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.38769507182444535,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1358,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.3971610674446132,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.133,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.41088534710317975,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1566,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.3918183093661922,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.141,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.3708984663987497,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1402,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.41579197132379475,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1629,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.3701651140764943,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.129,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.4061100588397826,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1331,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.4001656508592521,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1596,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.38461237909070534,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.142,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.39417104535731756,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1478,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.4270323997265909,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1417,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.37353604340569946,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1305,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.377341751068144,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1267,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.4020501165795556,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1482,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.38988959564712156,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1473,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.415993035859317,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.15,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.41939468262609775,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.156,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.35057500920432544,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1169,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.39434818856239856,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1434,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.4167557968688714,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1465,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.3912203268500574,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.137,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.38111035948703875,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1364,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.40069955460526036,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1434,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.40395702570817277,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1401,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.380803112950561,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1458,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.3733920615033602,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1513,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.3713219353912598,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1529,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.3883781593279937,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1458,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.36182690415736507,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1338,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.399023521143104,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1564,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.37808852781376256,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1393,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.3843133442279898,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1519,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.3584564349759436,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1243,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.38789081789331853,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1493,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.38499005445354484,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1327,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.40172827982097387,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1469,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.3998344024335528,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1605,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.37541950563938437,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1383,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.3732569675356344,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1583,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.34184604598405416,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1201,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.40498541181171094,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1564,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.36893923204226226,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1429,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.4035132112142162,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1484,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.3831655397823202,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.139,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.3946689669059379,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1293,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.362425774309565,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1145,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.37789884801224893,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1592,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.3795519678484875,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1497,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.3759827532495534,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.158,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.3776744436114995,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1595,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.36842245159566295,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1487,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.3636617983554039,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.136,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.3940448745804352,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1518,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.3899262301556109,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1497,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.37443581459783787,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1306,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.3741012911381219,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1556,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.3415484204512209,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1309,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.4008005187414092,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1654,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.3701168101381214,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1399,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.36654246420897735,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1346,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.4035271189532281,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1522,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.3698836347971586,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1254,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.39839065006743546,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1564,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.36090851129143436,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.1311,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.3801415429080757,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1559,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.4165127331306603,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1694,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.381677973940789,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1434,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.38060410413082796,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1271,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.39183254036934967,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1498,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.396753968349085,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1404,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.39656529033889526,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1499,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.37836599847307495,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1419,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.39875327707526304,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1486,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.3518793526309351,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1194,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.4047415072278516,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1575,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.4418867558307131,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1261,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.3948579191081847,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.146,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.36104647759548564,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1382,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.38490776969295226,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1461,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.3725887619711043,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1414,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.36453242183542484,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1453,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.37172831023084213,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.147,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.3630820795969288,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1393,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.4039985934773245,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1532,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.37031496186508944,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1291,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3988366443210275,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1329,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.3722806024457161,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1395,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.37045354475569187,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1411,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.3936632368855796,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1577,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.37167201846451314,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1317,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.39698370466677796,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1588,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.37726745791811495,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1481,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.38959216031592947,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1483,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.37371693353170715,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1391,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.3786494478315997,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.143,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.39402219769135105,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1652,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.36973411273429985,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1378,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.3679339106087277,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1325,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.3577711540388804,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1178,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.39717743483357215,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1465,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.3862490573595438,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1409,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.38633145274065334,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1539,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.3740800558332791,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1387,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.4035015548518203,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1402,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.417335752573794,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1558,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.37925890203540263,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1281,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.3994654407480251,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1435,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.3892829938192953,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1373,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.37391224443989424,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1358,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.3899203521512544,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.148,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.3527441026287905,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1331,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.3591770395272059,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.133,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.3750553603068423,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1403,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.3871839021184729,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1522,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.405735104083125,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.175,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.3378040593276621,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1279,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.373681119908766,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1627,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.36413117188383637,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.153,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.3871465901276519,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.153,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.377609510326934,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1432,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.36092402955223546,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.139,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.38661230953799264,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1484,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.37611863130076273,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1428,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.40161547526877833,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1557,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.37493192395783487,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1311,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.37467157363481574,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1397,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.3999671865535423,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1579,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.3820753584742572,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1417,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.3725818542766375,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1467,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.37024963441612696,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1435,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.3793008956571512,
+      "learning_rate": 0.0,
+      "loss": 0.155,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.3387654019531407e+18,
+      "train_loss": 0.4389778661534069,
+      "train_runtime": 55204.6024,
+      "train_samples_per_second": 2.862,
+      "train_steps_per_second": 0.022
+    }
+  ],
+  "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": 1.3387654019531407e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}