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": 7.755090006988808,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.3207,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.018291494588603,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.3329,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 7.8992279882213525,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.3156,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 7.76459611964643,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.3165,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 7.511865028556111,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.3173,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.230484920161975,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.315,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.808987040841214,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.2294,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.502257739890044,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.2828,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.733651226606316,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.1643,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.6089329088441797,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.1788,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.61479159452278,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.1824,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.2708470045455496,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.1134,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.597567016108322,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.1502,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.153106263190792,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.1431,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.4724773185652302,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.089,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.363427519463842,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.0792,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.76290865279589,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.0436,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.750068195510758,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.0727,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.1925313373014848,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 1.0238,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.7871114900096883,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 1.0246,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3473605056056988,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.0137,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.1789568183911903,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9764,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.344282330044553,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.9851,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.2273694667038069,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.9761,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.093418098836108,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.9657,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.0337613386968343,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.9723,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9991117538592656,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.9681,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0782307363385872,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.9796,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.894951917904586,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.9607,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.9613647157681753,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.9596,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.8406520027644168,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.9408,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9084886192178246,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.9314,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8218983512317202,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.9399,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8792808417980341,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.9518,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8927356701994327,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.9557,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8529732424319364,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.9216,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.9582388591066847,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.9514,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.8360798056827038,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.9322,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.7836817157447072,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.9354,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.8598372221734446,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.9217,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.8531829817342795,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.9336,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8332473062152559,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.9116,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.8095008680877583,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.9093,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7729366254933746,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.9053,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.9004079857333699,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.9212,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.7697473689429393,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.9199,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7894372762821114,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.9142,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.815495795117999,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.9044,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.823631997249554,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.897,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.8648003591369358,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.8899,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7863201574289026,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.8906,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.732784268945566,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.8972,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8044693573329423,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.9119,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7208501298178833,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.9062,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8148511149342993,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.8916,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7909692044876077,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.9061,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.8060548545785714,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.9263,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.761931427201729,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.9132,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.8488685006958424,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.9087,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8188259676118809,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.8815,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.905354265952662,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.9163,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.8822188246406978,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.8886,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.891060089091389,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.9377,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9441989622947415,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.9093,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.9391706029896636,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.9134,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9653311008384666,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.8845,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.8506770711515365,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.9102,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.0548140021688457,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.9002,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8986839692185339,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.9115,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9080041387845066,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.9131,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.9298245341901095,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.9044,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8604633103090714,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.9052,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.044697606688382,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.882,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.013502180870298,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.8796,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.9557410667286693,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.9004,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.1825166310243822,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.9199,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.9755570787053145,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.8879,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.1319322053384009,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.9009,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9292469399805244,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.8933,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.233381641206865,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.8793,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.8961256847130804,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.8981,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.9419586155826097,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.8763,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.068078657865175,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.8824,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.8348238965310619,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.8735,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.008274502645204,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.9051,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.3432143219867367,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.8663,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.8902653755412601,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.891,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.5043483114987728,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.9052,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.9901829252081649,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.901,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.8124862578486198,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.8885,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.153785197727016,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.8934,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9458759977736693,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.8983,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.8701425912910248,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.89,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.1011703310695178,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.8875,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.3542810177091558,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.8745,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0426681543736156,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.9071,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.7673541244029514,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.8853,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.3841540534586947,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.8601,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.559685762296352,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.9043,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.2324294291582003,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.8756,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.1618086053228804,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.8711,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.567635073773976,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.8617,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0229285893082252,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.8914,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 2.2849150223852512,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.8871,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.991568171523068,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.9147,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.9424860488009832,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.8897,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.89240575907554,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.8723,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.4635764726019362,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.8748,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.9768301054236805,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.8997,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.2612289108000818,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.8717,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 2.248164368292501,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.8963,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.7100031647344094,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.908,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.6798626935517917,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.9005,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 2.3570660504022043,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.9196,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.4106807419831409,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.8866,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.7740713934748116,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.8866,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.7955109723471008,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.901,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.2166426378546917,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.8671,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.6105129696412854,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.8839,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.4707000067930496,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.8634,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.4741736038157711,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.8835,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.471978923636948,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.8878,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.3356616379287374,
+      "learning_rate": 4e-05,
+      "loss": 0.8714,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.2301733163160102,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.8848,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.3000942454443234,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.8799,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.257012176551839,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.9028,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.8080932887558345,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.8986,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.0309660402899095,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.8269,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.8790392620833991,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.8773,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.1094417971603239,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.8659,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.8526442697247882,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.8553,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.1703100350078444,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.8797,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.9399820636939593,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.8641,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.1351852635776516,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.8612,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.250436816171775,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.892,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.905677045501483,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.8742,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.6511057135607838,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.8643,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.084373211815271,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.8746,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.4106460904010667,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.8971,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.1029631189984723,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.8777,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.3057400565929096,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.8776,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.059355144704799,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.8713,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.1926861504504882,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.8649,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.1475152006931095,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.8833,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.110540016861538,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.887,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.019002904388605,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.8662,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.015381520946677,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.892,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.1352033904918948,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.8844,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.9002834161574232,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.8666,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.120803731268676,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.9009,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.9823828357977181,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.9079,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.2345567810409523,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.8793,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.0057261594783857,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.8872,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.1956082313021497,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.8772,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.9945030411772742,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.8804,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.344911399754115,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.8925,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.9270350573190638,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.8722,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4584163165139932,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.8855,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.2421261982573473,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.8764,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.1459236916571658,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.8788,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.3090338033834574,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.876,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.080742874935174,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.9,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.963184188099128,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.8997,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.0297553807616535,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.8784,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.1629148534644156,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.8868,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.8601375561486628,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.8856,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.232128136101819,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.8875,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8330360677727551,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.8853,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.2928803962473079,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.8938,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.898393434227544,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.8941,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.235888310898347,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.8643,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.1482095326868202,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.8757,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.208590104362501,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.8887,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.2521337646361075,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.8826,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.9828893812344124,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.8841,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.1398280704911032,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.8488,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.1092820496270666,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.863,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.0209409107310259,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.8999,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.9207616709610813,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.88,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.0794495365338883,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.8608,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.9650823847479795,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.8816,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.0978285750946344,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.877,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.0100250257604024,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.8776,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.066104835539133,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.8788,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.1389108868447388,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.8986,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.7981617526034015,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.8747,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.912489835815976,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.8523,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.9257486019007773,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.8699,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.0444036540821073,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.8806,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7933167132221397,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.8886,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.024711932674448,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.8665,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.8169504493723225,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.8591,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.8725055866609405,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.8827,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.8029630669771547,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.8661,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.9707459055329561,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.8841,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.9022384114757741,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.8723,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.9412149977780648,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.8536,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.9466078697331242,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.8909,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.9310220710530019,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.8627,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.934634872730739,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.8926,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.1930398651793779,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.861,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.9148787498354868,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.8763,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 1.1352798790292609,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.8652,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.9798975955009194,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.8577,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 1.1087979107836914,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.8884,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.2501429229639107,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.8882,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 1.0315443391857555,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.8856,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.3423229987150709,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.8718,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 1.0985596373503594,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.8762,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.3065343195735013,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.8821,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.9914188928999979,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.8402,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.0679356265414772,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.8699,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.8981587374010555,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.8806,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.9233539921479532,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.8491,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.8090292425010747,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.8738,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.7135760334661649,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.8603,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.9723716397936014,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.8617,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.7532744068425842,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.8874,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.7604673091094274,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.8757,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 1.0241744628685632,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.8739,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.713108487324263,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.8768,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8969075127552222,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.8804,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.7757654932712001,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.871,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.7706073017711785,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.8727,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.7544945907532029,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.88,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.8649270750382847,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.8443,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.834177306939768,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.8587,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9571981698466445,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.8729,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.842993229581042,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.9026,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.9621775094245316,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.8877,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.8787151039759762,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.8826,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.7605456691424366,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.8493,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.8078065840163957,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.8765,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.8196489280515299,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.8717,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.8085207454465597,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.8819,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.8741797531268085,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.8666,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7656476997703653,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.8636,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.8043664517647391,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.8746,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.8527872591039067,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.8522,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.8136574149400297,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.8623,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.9292457609768098,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.8793,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.9183549523588894,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.8764,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6914343303174342,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.8584,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.8498503129335762,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.8381,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.8153630827727306,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.8645,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.7912196079149352,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.8589,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.9062229035294436,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.947,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 1.0666247576317482,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.703,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.9140820023619474,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.7049,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.043806509563789,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.6847,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.9581184500399132,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.7077,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.8736859873663831,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.6291,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.8557574638958139,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.6553,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.9508892579751134,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.7048,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.7669637135896662,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.6576,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.8398535276322447,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.6936,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.8958988046672952,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.6595,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.9244626810570157,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.7095,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.8194414124366086,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6762,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.8099390716903087,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.6636,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.8497607143024111,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.67,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.8057615112845727,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.6592,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.8626786856882459,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.6824,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.9616038744885705,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.7223,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.6475271574073478,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.6029,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.8494723674519057,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.6755,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.7903464059135161,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.6583,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.7702964158486125,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.6716,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 1.0028241591492864,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.6805,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.7793371642007065,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.681,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 1.030344049093498,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.7005,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.7995041695815587,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.6599,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.791149987196619,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.6599,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 1.1180716253680996,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.6776,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.7017449463883065,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.6572,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 1.0672822848688337,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.6673,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.9562823915546735,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.6602,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 1.0200388661185495,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.7017,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 35.32725653687703,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.7227,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 10.531556421728972,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.6897,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 2.2585475960590418,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.6605,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 1.1969284501280988,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.7056,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 1.9146287846804653,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.6943,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 1.8052263032469251,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.66,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 1.311628822137459,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.6838,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 210.63959807602342,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.6691,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 4.932488142576597,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.678,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 2.1492627749893902,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.7074,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 1.13787648299952,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.6998,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 1.503279616022418,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.6731,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 1.6824326081904775,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.7104,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 1.4229370025642523,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.7034,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 1.286001798323535,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.6946,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 1.440536739806033,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.6759,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 1.410718427796322,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.6953,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 1.1092165884846148,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.6954,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 1.2065804432099558,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.6752,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 1.0265100699784633,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.6787,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 1.0581603578435106,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.6538,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 1.026135554484098,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.7267,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.8075229217790061,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.6551,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.9118617674952505,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.6611,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7782869135202596,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.692,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 1.0604919848368284,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.7031,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.7734668041249066,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.669,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.8305224688207471,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.6762,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.8679534398464125,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.6898,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.7887961907629109,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.6627,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.8271010658653666,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.6801,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.8727435753513556,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.6766,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.8032092665942296,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.6661,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 15.772037218366881,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.9478,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 2.683978034431614,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.7226,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 1.239683320798963,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.6692,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.802177823534339,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.6776,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1487180128928844,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.6914,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 23.388759337004164,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6885,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 1.1768968781234077,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.7108,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.7639753211516357,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.681,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.9474001733545495,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.6658,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 1.047047356622185,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.6652,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.9600212198604301,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.6809,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.8400416655368387,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.6887,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.973586064531452,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.6811,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.8075081819684632,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.6765,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.731977628746432,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.6469,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.861639998665284,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.7205,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.7743163852721575,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.6363,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.8451886825246084,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.6709,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.8261836953986098,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.6892,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.752681662192974,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.7026,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.7791672913586524,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.6413,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.6719397446439856,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.6504,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.7257627138354701,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.6765,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.6416182532368547,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.6657,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.7536005844853645,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.6723,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.6689770705075295,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.6811,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.7639452312689767,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.7041,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.7186145867905621,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.6534,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.8289908615094262,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.6847,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.7489892888315172,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.6702,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.9529583393324838,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.7083,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.7916866860388665,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.6768,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7955099867500531,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.6936,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.6856103982083724,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.6862,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.7403455869371882,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.6895,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6673341800524946,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.6742,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.7664924618702114,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6702,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.7816859026039857,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.6703,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6591078612791326,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.6686,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.7428547376247264,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.6885,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.7192901652346437,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.6879,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.6772771650725272,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.6548,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.7461358472529636,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.6724,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6928052008469033,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.6801,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.7165085191280098,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.6589,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.7751221460012699,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.6894,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.7163483737368489,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.6762,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.7202709245540018,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.6746,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.8331036898779418,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.707,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.8251123333219025,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.6903,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.7830792847786582,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.6891,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.7915699647618387,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.6643,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.7381632698352191,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.7048,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.8968670174897585,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.6738,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7337746230482431,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.6957,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 1.0229097827753395,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.6937,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.8623988960779362,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.6565,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.8653087922709795,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.6968,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.9063322270008759,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.698,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.8157066361686913,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.6435,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6972306335263415,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6709,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.8729547419123075,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.693,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6757762567648384,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.6774,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.8102823565792864,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.6557,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.7303027664856511,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.6662,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.8041753390323934,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.7314,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.722988265963281,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.6964,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.6517660632084685,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.7009,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6956772125570081,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.6779,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6793028492903811,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.6865,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6671712662412815,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.6823,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6403648917372665,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.6786,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.7312080470219897,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.6593,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.8747626611723297,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.6951,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.7877319093288477,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.7038,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.8032411753000075,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.665,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6916424530292041,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.6613,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.699932554478291,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.7004,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.6579206933477294,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.6818,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6462316517464401,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.6779,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.7761874587742155,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.6768,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.667347786173353,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.7198,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.7830186244841448,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.7011,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7034697264987454,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.6487,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.7020415608119113,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.6953,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.7693215122061521,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.6961,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.6235317916271209,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.6639,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6639464302333213,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.659,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6533701498314701,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.6978,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6621454665000295,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.6779,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6674830180418145,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.6896,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.7271898845262201,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.6808,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6356224489857635,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.648,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.7283858140740971,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6843,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6811942349220751,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.6428,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.7493081288503299,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.695,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.7725803680447756,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.6862,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.7777917045735102,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.6683,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.6781585464617792,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.6619,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.8290878105323721,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.7097,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.7635040627666819,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.6708,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.7449819305201636,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.6799,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.6368455332560411,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.6501,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.7184992548245759,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.6736,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.7041942824057497,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.7051,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.7086751311882782,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.6735,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.698222181497206,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.6879,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6820284574337219,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.6971,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6409956816756146,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.6646,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.7251583560713226,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.705,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6482841942179293,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.6526,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.7832598465941244,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.6817,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6949537200696845,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.6815,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.7408543864848465,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.7026,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.6032467601997837,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.6538,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.7569598109604669,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.6806,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.5988641031555803,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.6744,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.8086972364014997,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.6626,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.6503528059697415,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.6699,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.7695990681772575,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.705,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.7215371309187516,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.7071,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.6500768388597481,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.6332,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.7590759573126025,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.6941,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.6856383053920374,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.6825,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.7073941257747937,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.6449,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.8000635559273953,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.6753,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.7306809967562177,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.7105,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.8401634334160192,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.6681,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.8514682040852969,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.7096,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.7683477167319845,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.6583,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.8215737457818667,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.6685,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6583156124826272,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.6983,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.9812814996196219,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.6591,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.7076325724311532,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.6631,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.7811550833750038,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.6947,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.8380389096993153,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.6813,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6989627131374548,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.6763,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.7893728797309386,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.7108,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.7003200960812495,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.69,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.6613282188422246,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.6674,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.7976823923495134,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6531,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.6551128890675277,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.6678,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.7473050462244318,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.6984,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.7677492798674073,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.6836,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.8071547887189015,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.6665,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.7746778367213597,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.6706,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.7488703054091328,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.6939,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.7916002797739233,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.676,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.7186290732902395,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.6854,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.7718479505197589,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.6691,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.7863824850952955,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.6415,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.8393655385575794,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.6816,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.7602337609168961,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.6614,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.9497552754267619,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.6541,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.7608201923437865,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.6739,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.850269876197071,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.6864,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.7681202292648448,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.6676,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.8060241426625063,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.7116,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.7052587235819541,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.6589,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.7423554263959942,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.6837,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.7553130878116127,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.6728,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.7043954518495483,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.7029,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7368775543412529,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.65,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6937701453628914,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.6825,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.6645466086924475,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.7012,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.724789991300859,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.6979,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.825082570295718,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.6683,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6892173059761099,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.6519,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.8620143100254236,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.6543,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.784015984954719,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.6818,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.7308402594828607,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.6433,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.841751176682197,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.6929,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.790174322780695,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.6758,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6784047770692518,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.6655,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.7249586342648147,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.7041,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.6709164257155579,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.6731,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.7376830712134653,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.6754,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.7665718818432286,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6804,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.6781969077187231,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.6606,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.7877634203777095,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.6881,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.6238242652464305,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6696,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.7662416272257999,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.6621,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.7464444258498469,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.7023,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 1.0074280544053102,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4449,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 1.4718413413938582,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4169,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 1.019099891313021,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4382,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.8583272745916093,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.4032,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.8746637498733276,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.4458,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.7160541024448294,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4236,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.7391552403224677,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4247,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.744303590419504,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4011,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7766848816291504,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4292,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.6990635549007931,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.407,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.7262684519574482,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4441,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.7466434659698685,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4063,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.8465819987142408,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4291,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.7519963714559725,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4513,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.789368961734284,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3878,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.6559410130188464,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4344,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.8059004455469195,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4107,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.68034331939964,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4256,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.7574468329444276,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.4641,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6348215348469994,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3774,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.6915033448802093,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4161,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.5868342166080441,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3821,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.6940413479671568,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.4276,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.7103590718752212,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4245,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.6458403393222304,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4153,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.619534698981062,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4166,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.7202125470159692,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.428,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6541086279019175,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4268,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.6167245677500872,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4431,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.5835903107668045,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3896,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.5890495430158554,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4007,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.6860797920888488,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.4368,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5817672246064647,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4088,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.6280582043043436,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4335,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.5979429637970644,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3818,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.6089962897337429,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4193,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.5652935103226653,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.4019,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.6790458631571665,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4545,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.6345531275034398,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4437,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.5807956387983352,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4047,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5990696372096423,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.4003,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.688740470023339,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.407,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.603847950033631,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4356,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.653888739147002,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4198,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.7128114784469958,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4301,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.6305212050643593,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.4093,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.642325903783674,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4154,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.6663893555530531,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.394,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.6504834197626443,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.4291,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.667929234385205,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.4202,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.6761290393535875,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4203,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.6730670127247862,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.4308,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.6170983769224869,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4003,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.6157595189471586,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4272,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.6288963391724437,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.4548,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.6117842355596351,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.391,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.6644454610639483,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4399,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.6151598746901418,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.4245,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.605243732651479,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.4051,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.6780581873746898,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4413,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.6191960391603726,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4429,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.7100668821444378,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.439,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.6243522764307082,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4446,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.6427322221288565,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4043,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.6889691008488384,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.453,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.6391919620872507,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.4146,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.6500538351585823,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4188,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.6789748509693094,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4201,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.6979991865141856,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4272,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.6050616402246983,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4258,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.7062064840915108,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4425,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.7321855017550476,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.437,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5851617922208064,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.41,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.6458003997894223,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4089,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.6408479498402145,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.4181,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5567135479982934,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3947,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.6940121146809987,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4503,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.6228182421488494,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.4423,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.5623209138231197,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4002,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.630643973548174,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4261,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5983625367535629,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4379,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.6005569918647122,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4406,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.6260550864676598,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4232,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.6168160288529404,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.446,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.6261920196688817,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.43,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.6111603999289409,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3996,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.6553481828190697,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.4439,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.6284440791969241,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4276,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.6326560919471567,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.4363,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5862408977886105,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.382,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.6050642891832898,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.4401,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.6263936733016838,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.4294,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.6355998678026572,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4452,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.5948453800722175,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4292,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.6068533223585679,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4051,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.6393803614599016,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4378,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.6411031193149267,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.4538,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.6402356564192014,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4176,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5832587226034909,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4125,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.6146642591377574,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.4441,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5645611299090646,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.4005,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.6205580765410624,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4569,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.6052851025689058,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4149,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.6796127410967109,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4399,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5454433794973353,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4083,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.6982788957747564,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4415,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.6575768835936523,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4307,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.6013069248976044,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4205,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.6910088514318089,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.427,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5950338207557938,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.42,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.6113797199448774,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4177,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.6457219781050634,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4293,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5504806588824669,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.419,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.6419734195378004,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4281,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5751450062944389,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4155,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5844072355314723,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4128,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.6429904965665404,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4559,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5517029302704917,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3846,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.583975990015445,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4295,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5710486469963065,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4401,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5903817383713124,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4141,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5947096774645498,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4232,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5954493534851031,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4325,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5653642368034567,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4185,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.6332431090381087,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4049,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5665159963636622,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4247,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.590778142286487,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.437,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5652141034749675,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4148,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.6071880159141638,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4483,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5732844734417827,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4091,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.5910957983171254,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4139,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5738747133014489,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4261,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.6010675216128687,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4051,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5952788683919377,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4319,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.579149939934123,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4035,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.6196817271946797,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.4329,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5593411781725489,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.4463,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.5776911470471904,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4263,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.6162524032961144,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4194,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.6017336116385316,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.4349,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5524934931848642,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4017,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.6098978203504576,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4209,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5873139091992933,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4286,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5892093804306463,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3995,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.6156664229428928,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4136,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.5600527491265884,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4256,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.6098236775266107,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.4319,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.6665504874306813,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4506,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5716255327591399,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4232,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5234982042334299,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.388,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.5752346804096994,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4256,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.6564691335939125,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4593,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.5722702862765491,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.419,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5566324482807752,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4065,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.6115848178094498,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4555,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.5890548829527964,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4077,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.5821388771098703,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4016,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.590954861602569,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4197,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5425052634542877,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.4063,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.6381274967718513,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4569,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5432116994137807,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4067,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.5702998267165831,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.4367,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5703168188560153,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.4297,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5901256505340442,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.4341,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.565401957164755,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4131,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.6257378105880297,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.4329,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5733814180833737,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4318,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.559952528373068,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3993,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.6606387948824293,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4428,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5957665386404872,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.4261,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.5682008939022933,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4106,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5853588867409373,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4174,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.5949412688473726,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4224,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5749449913997887,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.4217,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.576824920082504,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4266,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.6363654887520214,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.4415,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.5340637131645248,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4149,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.6099695344702116,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4409,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5407054848844368,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3979,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5785329671476686,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4315,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5872008301981636,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.4383,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5713608006214975,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4138,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5695971714896478,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4293,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5149467540185659,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.4004,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.6151647946294759,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4195,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5950969619032686,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4554,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.49804799694283935,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3521,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.6241972682344905,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4792,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5794883412820785,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4019,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5699895313416419,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4463,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5808678056003451,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4246,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.6067138860682412,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4418,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5250438197176253,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3786,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.5937066169692657,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.431,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.6367200102127571,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4206,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.5602749438300382,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4328,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.6623424202558515,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.4475,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.6175991180819624,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.4272,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5615804148238344,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.4202,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.5620490327184613,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.4059,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5926348551902668,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4232,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.6234337892502043,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.4464,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.6081193513577176,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.4164,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.6089197307399097,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4471,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.5833542992360782,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4108,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.6138850283010113,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4319,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.6028219117097514,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.416,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5958069542551349,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.428,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5573587082308129,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3996,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.5976750466697138,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4336,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5749220533059143,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.4233,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.5726747921203943,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4325,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5566485563333008,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4026,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.6977244598237271,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4787,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5359179546524533,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3937,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.6058470632030009,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.409,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5867742489618623,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4362,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5668989375614033,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.4163,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.6211949144857275,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4408,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5986439273206641,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4349,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.5398277350835909,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3946,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5895179805371377,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4355,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5681871206026755,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.422,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.6025999094455714,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4406,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5236633040975268,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3877,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.5738067943243754,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4204,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.5981294558204013,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4272,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.535820282867727,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3982,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.580657401513468,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3966,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.60521029643708,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.4527,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.6143774551704372,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4306,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5552245645590871,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.413,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5824216618401635,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4358,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.5977064780523549,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.432,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.5554455912709686,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.4103,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.5325525513742276,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3921,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5785334105295219,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4146,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.6340031052083491,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.4399,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.5821062262665906,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4056,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.5467682819723629,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.4136,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.5709666393144881,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4063,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.6109955655865187,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4247,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.5581352051020211,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.428,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.5535977933037731,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.394,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.6191790198811536,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4415,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.5727028974564234,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.441,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.6785002459435414,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.3771,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.8010356746358883,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.212,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.5622506177919311,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2155,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.6489371780491717,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2267,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.866037334416392,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2144,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.8032109860413128,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2129,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.6591956039166434,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2378,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.5376987644887444,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2135,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.5917791390789613,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2143,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.6396972155910883,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2248,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.556008009555202,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2033,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.5273032568650774,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2202,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.5238864633252169,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1981,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5842754897969351,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2011,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.5498077682071437,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1866,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.5833372504593715,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2128,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.5281429921094127,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2113,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.49517048887894755,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2079,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.5181184492632128,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.198,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.5645472993410547,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2299,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.5156374427029391,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2026,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.5046023590410988,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1917,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.5289589164232594,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2088,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.4903470155448499,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1922,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.5375875941685155,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2115,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.5210759464500875,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.1964,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.536024185802536,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2241,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.49785931933818506,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2066,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.5193293198237338,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2009,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.5229737786736439,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2277,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.4956737617443919,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2067,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.47041064577389374,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1841,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.5272387139883603,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2164,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.542307750294423,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2282,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4892445852890936,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1953,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.5088456617490071,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1915,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5458024831021207,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2212,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.49846731664765986,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1983,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.5168257525501019,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2006,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.4828184991160967,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1938,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.5267738121111825,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2226,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.5139942330344391,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2032,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.512208982146257,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2202,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.5770230785996492,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.242,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.5158093222684876,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2068,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.5146391202445293,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2018,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.5047010223727725,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2155,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.5043268055388915,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1899,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.5185117411380034,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1838,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.5414811898741477,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2321,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4984659844389358,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2017,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.478601316990032,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1897,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.478100787271925,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.1914,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.5013743010419418,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.192,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.49712730769036173,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2014,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.48097547509310923,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1931,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.5192658665160268,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2097,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.531806959874807,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2117,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.4742906177449081,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1956,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.5466174475792197,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.216,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.5118280282174995,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2154,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.5133008345159039,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2128,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.5539993197797799,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2182,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.497018787148489,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2136,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.5193906732522088,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2064,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.5237699879442144,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2198,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.4950293754076955,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1971,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.5105790228641045,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.209,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.48074448383143487,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1926,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.4894333272588077,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2056,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.4751196033382256,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2026,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.4935227447165871,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2153,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.512291556410902,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.21,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.5051874898176544,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2086,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.49686515104824996,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2049,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.5044338358555817,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1919,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.4936702369376674,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2078,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.4846176539605031,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1923,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.48532082419454264,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1916,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.5085847831325326,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2152,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.48998147371489364,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1985,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.5318975047797627,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2175,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.48398849810601047,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1837,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.48875339550938707,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2027,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.5196124395878404,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2036,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.5351790612883378,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2268,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.4715525520825065,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1868,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.4882146530454133,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2126,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.5046255481777153,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.196,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.5120519625303243,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.22,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.5012423342033623,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.2107,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.48782248524348903,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2065,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.4606508607100068,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.1859,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.5039444278107879,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2103,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.48539376318579014,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2044,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.48418448430000555,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1858,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.4992259416840778,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2143,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.47236746808678653,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1904,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.5051941185414137,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2071,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.510990719900377,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2081,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.4917549791013881,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1923,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.4952352506298071,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1997,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.4753765701400376,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1796,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.5125691387493422,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2182,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.4924604870852991,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1986,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.49964789399212545,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1982,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.5030379368241291,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.5040575836982542,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2076,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.505052155787338,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1899,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.5092517093362857,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2098,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.525669937027973,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.2295,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.5011244133427635,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1928,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.4861104103438396,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1716,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.5074591546002547,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1954,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.48276518992857276,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1872,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.5082239101501291,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2092,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.5096109179087047,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.196,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.5031896811311825,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2022,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.5101109546373928,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2246,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.48921162665069606,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2005,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.49774229734321696,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.4969219433587677,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2113,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4482503514551019,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1786,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.5307641149668371,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2184,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.5171461007609547,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2186,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.504234559012798,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2032,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.49576602128863184,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1903,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.45616426098890805,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.189,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4702080056194264,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1958,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.4964540586979825,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2034,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.4975389252499916,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1942,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.4964291340097882,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2047,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.49058359727984086,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1924,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.48965568543591964,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1961,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.5094106945906227,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2186,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.47994550140733083,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2045,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.4924224151763942,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2137,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.4606288805937345,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1927,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.5247741740960737,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2209,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.47517535981704007,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.186,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4950100887532531,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1859,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.5113563864579502,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2256,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.49931951901380195,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2024,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.4917772750589504,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1972,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.5102075252094873,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2081,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.46639381870783814,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1813,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.48663356578741773,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2022,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.4721278859710191,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1985,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.5177375427833854,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2168,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.4932835826881154,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.187,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.5399430322857165,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2439,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.46177644508679205,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1744,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.5107157110495532,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2032,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.5035719938913953,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1964,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.5326500211284018,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.232,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.4796199423324951,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1935,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.5166590453525998,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2152,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.48045033247043883,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1932,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.4906993660547527,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2068,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.4964819129976639,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2033,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.48085024508428037,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1939,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.48198877710299787,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1895,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.5041360780117257,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2143,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.49931102837131747,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2041,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.49297870737340654,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1886,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.4818938799287047,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.203,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.47200663195505793,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1944,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.4849703709597862,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1905,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.48408264042389276,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2041,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.5019006143732267,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2041,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.4928252721500139,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2119,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.4709524234271484,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1978,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.4748138427723408,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1997,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.468415979069043,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.183,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.4798416286674506,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1856,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4984294670665285,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2081,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.46897166529359813,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1924,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.5110928441990952,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2099,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.5194146225262396,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2068,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.4746522783121807,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2004,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.49049560688094546,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2104,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.4666790091912798,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1917,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4995943109566573,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.213,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.4881431400583031,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1963,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.47229906030998864,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1929,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.4923333577926189,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2075,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.48381880363591545,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1903,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4971641103056941,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1961,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.48114949152295755,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1873,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.48338246346766317,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1891,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.47479301542148356,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.176,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.4951808561299146,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2104,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.4945056128480562,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2081,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.49140609924928075,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1965,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.4916732144991452,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1805,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4529488978664983,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1788,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.5024832336970184,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.231,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.48104749547559916,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2003,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.4781802239986071,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1811,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.5096129115668606,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2036,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.494876496210906,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1953,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4754002243838514,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1954,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.4769614810690057,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2004,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.4944975035224176,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.19,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.48105670833520703,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2046,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.4796759545172689,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1927,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.48470197551150956,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1927,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.47541137676539286,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1803,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.4850848606751684,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1981,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.4502070757579807,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1884,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.5207992162364101,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2074,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.4517451545813081,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.165,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.5001571606393458,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.198,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.4620936905968506,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1903,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.4895385156952171,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2111,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.48939451718245286,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1973,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.47701005087868137,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1934,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.483257802144118,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2066,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.461950255390057,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1906,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.4446670243733804,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.1887,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.473578347891457,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2154,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.4734877070824314,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2119,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4622941542175381,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1894,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.47346141052223506,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1965,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.4853252858683708,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2087,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.4848028679584145,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1949,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.46553559900891833,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1809,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.4961349476153128,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1943,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4892378385746828,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.177,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.5005182921522481,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1941,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.5062374154757002,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2121,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.4718144108505057,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1878,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.47660439602385357,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1819,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.5127861103161294,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2034,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.45550922154473955,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1906,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.4767515487635344,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.198,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.492330791079021,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1949,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.46778294214603855,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1978,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.4667919246239407,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1869,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.48778519433238204,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1963,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.499747347669269,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.209,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.46666961952326014,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1867,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.49153772881493873,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2119,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.464126865559559,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1956,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.4838058559669622,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1991,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.476044064477332,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1773,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.5417500425276714,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2282,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.49886012194762935,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.148,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.5527265257195927,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0922,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.5333944328055538,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1047,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.46822953643356524,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1055,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.3858543894400976,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.0888,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.38112771971290427,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1045,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.39668391539999526,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1002,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.44142189468861365,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0984,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.49210477733253905,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0939,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.5162160112810504,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0903,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.502563799336053,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.087,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.5444444321232518,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1056,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.45309723641127153,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0851,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.47576853847847095,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1079,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.4095001362646933,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0963,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3592007306376557,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0864,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.38874047222514296,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0997,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.38749715103936333,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0961,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.38761228663168323,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0982,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.37671299955767495,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1018,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.3924631708627024,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1045,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.34652953732302105,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.0796,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.36312445986984104,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0894,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.3553493292339896,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.0914,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.3732472930726841,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0903,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.3837019224177107,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1016,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.3900364009663447,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0983,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.38705900470872584,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0995,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.3824910833306379,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0947,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.386275524402454,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0918,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.4098026088194475,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.0939,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.4198079211681232,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0944,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.3851029547258642,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0882,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.3666184971937843,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0931,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.39184720776016085,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1027,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.3791667899662593,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1046,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.36701982998443206,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0866,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.37937186698499914,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0958,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.3590691936878877,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.0856,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.37884045536134675,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0984,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3902009164365246,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1057,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.37770723407076523,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0931,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.38384442137246555,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.0951,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.3696646750148856,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0904,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.36561048705467425,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.0854,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.3764802612415719,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0876,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.40089006204941424,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0963,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.3802774503695267,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0866,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.3862676879612491,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.0959,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.3528885771339724,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0833,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.4057102967664334,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1071,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.346483780269309,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.0768,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.38055635253079373,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0922,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.38757671184901993,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.0991,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.338268416136534,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.0717,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.3918842967601419,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.098,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.3689876330071638,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.0848,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.37953446200235635,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1016,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.36718758964882986,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0933,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.3572208090782958,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0856,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.3896872501613211,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1125,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.3535272679470964,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0843,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.3676162366630153,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.092,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.3824100216640205,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0937,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.3720042857297976,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.09,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.33148977054453377,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0825,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.35105587157430357,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0796,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.3835031523253539,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1046,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.3756502925522997,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0937,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.37259310255485795,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0955,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.3565429253886779,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.0847,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.3800342179119921,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0944,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.34155721777038234,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0866,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.3915089663609232,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1049,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.3581699793684004,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.0953,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.3634872355007645,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0886,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.3721771235598181,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.0927,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.3642954697588933,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0797,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.3585011757514686,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.0805,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.3894734804733319,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1025,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.35976541145358126,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0896,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.371734561772852,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0874,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.34656440620053863,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0776,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.3914654600881387,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.0983,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.3728002848211544,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.096,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.3501623772395676,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0899,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.3630453984478295,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0914,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.39340689411137897,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.0901,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.3725953156198814,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0885,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.3848972457273557,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.0976,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.3798690737038849,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0959,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.35836391381498756,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0856,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.36644929880753824,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.091,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.38454619312492944,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.0876,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.3844998891032009,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0896,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.38168010445157974,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.088,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.3806193224203551,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0966,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.3744513142208306,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0969,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.3682987143932151,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0929,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.3575325242655614,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0802,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.374039867933358,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.0874,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.3958426614833818,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0922,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.3647970497112845,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.0871,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.36054676728424556,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0938,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.3761604677320592,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0927,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.3675470219749746,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0935,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.37432552189834084,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.0885,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.3617647385403568,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0948,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.3740568154264901,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0905,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.3588800754219211,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.088,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.3647805382677103,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0908,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.3723711044563446,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.091,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.37086502006763555,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0947,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.38014278448982103,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0964,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.35914359419995284,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.0917,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.3996499770257534,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1091,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.35589285411746907,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.0856,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.36984334658298723,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.0958,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.37092474468880937,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.0918,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.3637340698523622,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0875,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.38666088012933264,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0945,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.3818337206382421,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0971,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.35678439004048024,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0865,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.36721887158583005,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0796,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.3742207305969553,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0878,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.381358829772163,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.0927,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.3752624412914683,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0881,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.35725356551710896,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0802,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.3621683814393072,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0914,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.3740004455942084,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1018,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.3877526859311579,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.096,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.3730844118053592,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0888,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.35502461356450243,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0844,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.36364110560857077,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0869,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.3804554877014479,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1038,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.37371544512755434,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0872,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.38961002353688806,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0925,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.3507877430238908,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0784,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.35227279880981294,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0848,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.3965838553546266,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0903,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.3632840206641272,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.085,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.3912633044307653,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1005,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.3550003825871426,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.0735,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.38667859002938004,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.0985,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.3690957117478674,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0821,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.37607373512917414,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0923,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.3691790098372969,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0892,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.36787838065229905,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0909,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.3694724572653701,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0883,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.3729451335876711,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0834,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.37038339323415326,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0884,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.35961729946728105,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.087,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.39077168790263606,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0996,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.3583550018134953,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.084,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.34875122026271477,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.0801,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.37660269145332465,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0992,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.38165502609391216,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0986,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.3649198266893083,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0899,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.37830285575454325,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.088,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.3965063818387798,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0931,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.37177408138342394,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0952,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.3415398735408047,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.078,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.39421492150423143,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1106,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.37375910437789095,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0947,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.3337622336528515,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.0858,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.34788152774741077,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0824,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.35907257888496663,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.085,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.38796354993498255,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1011,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.3639831043855283,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0909,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.36784135861955797,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0866,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.3673014483296008,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0931,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.35308527236837595,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0847,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.3721460883856563,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0895,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.3688684828396639,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0893,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.3559295001253499,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0859,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.3723143074249689,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.0857,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.35758676131598427,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.0769,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.3936257877368265,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.1,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.34556089881155,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0808,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.3760056390608979,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.0906,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.38756179381838934,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1004,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.3426722769132747,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.0814,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.3807100054681576,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0986,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.36296735680099157,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0774,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.3866710055968179,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0998,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.3601133931606612,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0876,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.37702374530662963,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.099,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.3528469267104204,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0809,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.3847607991427551,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.0933,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.3674500789913031,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0825,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.3817336594862644,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.0907,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.37820126132626475,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0969,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.36530446909296693,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0884,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.38367606736652005,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0964,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.3699279171309375,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0949,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.3640349746097849,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0858,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.3552455012851595,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.0816,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.3798245397537096,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0899,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.37471323355470326,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.0875,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.3613214804844306,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0892,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.35882323834649144,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0819,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.3523831167740819,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.0823,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.39195530496676884,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1068,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.3680247575433471,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.0829,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.35500828454686856,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.0802,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.36332304288864564,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0874,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.3730834534065188,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1031,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.38469624927348245,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.0934,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.3684499466915605,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0846,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.38655496722966265,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1052,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.35943438210710016,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0793,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.3756129219079196,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0946,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.37995316215154273,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0934,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.3753596014243808,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0928,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.35684189527390175,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0868,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.372521744227299,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0975,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.3661356328180427,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0824,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.3635686558980601,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0788,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.3818457743301352,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0958,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.3689995274664308,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0879,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.37607807027148354,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0905,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.3932350965449307,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.0916,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.35582784293651143,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0767,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.37474699823136753,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.086,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.37338973458653874,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0862,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.37550311413388937,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0954,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.37067793882317135,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0957,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.3725149025628211,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0891,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.37648482129839544,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0983,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.3341748020711108,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.0826,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.37580750404901603,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0973,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.37973213962685876,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0909,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.35555345733917154,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0843,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.3745031532410309,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0961,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.405441210522081,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0989,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.3822198669862067,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0943,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.36644604340112197,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0881,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.355196128713554,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0836,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.39042061157706004,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1011,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.36270147561633204,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0822,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.3867165528457715,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0955,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.369064327348974,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0869,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.36075640693865896,
+      "learning_rate": 0.0,
+      "loss": 0.0895,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.1129855183851356e+18,
+      "train_loss": 0.46361072231235545,
+      "train_runtime": 41853.6475,
+      "train_samples_per_second": 3.775,
+      "train_steps_per_second": 0.029
+    }
+  ],
+  "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": 1.1129855183851356e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}