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": 7.848413499532932,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.231,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.384428864529255,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.2829,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.138636972855503,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.2475,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 7.791532489692888,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.2386,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 7.680868177179875,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.224,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.6606843732791,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.2505,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.0311050560099515,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.2114,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.336339133870769,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.1635,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.3817494350275275,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.1317,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.9692145506369623,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.1243,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.46286692890559,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.057,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.162371436025839,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.0573,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.35261676336337,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.0482,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.250195411615673,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.0512,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.6978138534696448,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9922,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.9845688181850845,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.0073,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.9014385402413616,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.9738,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.6097779402226853,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.9599,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.1285238130564066,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.9529,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.7930479163936002,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.9337,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.6229906535410197,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.904,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.4943271708770005,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.9067,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.8164825079519535,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.9291,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.5157434777647425,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8881,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.259899957119417,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.9068,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.2322153787127192,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.8784,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.1876936166101164,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.8801,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0925864719397975,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.8878,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.0111167534802032,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.8719,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.9164103784797052,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.8639,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9595508008135039,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.8525,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9549761826597372,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.8833,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.8779916305807526,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.8291,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8664904710095015,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.8424,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.825563579387797,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.8701,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.847806229349165,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.8441,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8127341577742556,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.813,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.8054821844490834,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.8451,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8333945300000299,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.8103,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.8625532118357768,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.827,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7738727023868198,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8141,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8834696694283957,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.8226,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.8225814068969012,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.8331,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.8130909429665819,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.8146,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6926207283641804,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.8167,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.8320873805244731,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.8018,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.8116650034505308,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.8348,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.8934491407212571,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.8214,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7922701491856394,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.8471,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.9062883480147156,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.807,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8042083877847466,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.8139,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.903797359003871,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.8272,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.9268669114876544,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.8023,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7478848165016431,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.806,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.9817158667010113,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.8064,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7327392223679243,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.7794,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.9128317767960545,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.8185,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 1.0590235487823265,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.7973,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.821600171808846,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.8061,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.1382338599926922,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.8096,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.0275365866817294,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.8072,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.0038983658822096,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.8171,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.9026958900745933,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.7935,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8892217537373631,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.8411,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.9141068531783929,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.8028,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7802162688485557,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.7826,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.9113161828710475,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.7918,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6945964498479016,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.8086,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8105903390683368,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7961,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.8018411532272601,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.8015,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.9136617497511397,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.8319,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8917770725736499,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.8089,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.9640569962448169,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.8169,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.9430741512476795,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.7714,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.8197093351782727,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.7773,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.06141430365222,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.8048,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8444727023968084,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.7781,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.9676701718011008,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.8137,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8378129972401226,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.8034,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.8764704033025478,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.7923,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.878238394249706,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.8032,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.1140166243885765,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.7946,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9102274723422996,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.8004,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.1055039760512684,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.7864,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.9087146563743586,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.7965,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.0123073653842343,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.7826,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.8304312235827485,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.8153,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.1643937324383613,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.8057,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.9244773163719588,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.7596,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.03917021604227,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.8162,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.0172036997867147,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.7802,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9314821111783246,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.8055,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.1451762017675364,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.7941,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.7848758444133743,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.7789,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.9426937856671089,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.7926,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0776215239576459,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.8036,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.7909769406346936,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.7698,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.146692663065015,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.824,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9018676045646569,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.8089,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.913935044408485,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.7984,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.9945175562997195,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.8088,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.8460888810839569,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.7956,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.9344986508607082,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.7822,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.9285742816095771,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.7789,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.8050165036983142,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.8071,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.9924929113971548,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.7795,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.8372469713340843,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.7943,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8356069116014467,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.7825,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.0442425500843182,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.7912,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.7977376639287372,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.7972,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.0222110585523092,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.7964,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.069728435010061,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.7955,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.7737784870815244,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.7934,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.0892270713304737,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.7864,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7778001220124636,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.7792,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9483019228132739,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.8112,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.9587681858200601,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.7875,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.8248683170110498,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.7828,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.0250510515694953,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.7857,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9530193582012493,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.8034,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.920909780943089,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.8011,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.9383156004689377,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.7925,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.9358247680787399,
+      "learning_rate": 4e-05,
+      "loss": 0.8079,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8902409541339763,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.8175,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.9807050332824817,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.7965,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.901561449275569,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.7977,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.7868919702944462,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.7963,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.9675601966216789,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.8101,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.7828637639811806,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.7916,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.7021528738328413,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.7925,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7981866608461597,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.8017,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7922923445494998,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.7853,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.7257878442412199,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.8082,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.6690792224287134,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.7923,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.847691165689436,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.8002,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.7481865526795616,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.7829,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.7717399271829327,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.8068,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.8492538231564778,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.8003,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8615327512067025,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.8163,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.8100820324226309,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.7993,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.9314960566632995,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.8085,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.8700279819367327,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.8029,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.8198652268053838,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.7825,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.8557053679560871,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.8256,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.6749012066117258,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.7816,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.9170592948828062,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.8203,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.7199901012732723,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.7967,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.8882841597393192,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.7865,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8186548243620234,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.7575,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.7535467807413083,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.7954,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.9272414149471546,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.8067,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.8226373664869655,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.809,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7620099713575147,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.7805,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.7641044054350895,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.7973,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.8838983853271608,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.7855,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.7706370112212982,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.7723,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7596383087979208,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.7738,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7461183934094657,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.8005,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.695732517954991,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.7957,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.6790585006561877,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.761,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6828796410202981,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.7953,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6670930015875425,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.8056,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.7230271326317966,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.8087,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.7034421111965145,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.823,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.7840932299020137,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.7873,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6820496391428105,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.8088,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.7199904477327841,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.768,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.7408402077903476,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.786,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.7843263909038984,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.7998,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.720814823066331,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.7925,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.7773275259607713,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.8233,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6691350373232697,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.7945,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.7784804486296635,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.7838,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.6471546743936819,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.7902,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.7092913673860758,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.8045,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6644456301346855,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.7763,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.7358586647523393,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.7941,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.7521075642819537,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.7854,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.6871815203332153,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.7835,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.7825690872853278,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.8017,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.637315434692007,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.7682,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.6202535705938703,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.7959,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7720064735686257,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.7944,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.6642376271508272,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.7795,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.8123499932563016,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.7955,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.6165455732145961,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.7602,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6739251606564804,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.8035,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6652646064300427,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.7652,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.6514152492892963,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.7864,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.802404211059515,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.7715,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6596945746036506,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.7762,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7163110606302553,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.8088,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.7772356752153994,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.7726,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6532087742609586,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.785,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.6891976141734723,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.7853,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6824107084372711,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.7919,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6776036100373125,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.7858,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.7098277057064031,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.8044,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.646183845365591,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.7563,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.7168917683374341,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.7774,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6437757181723707,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.7906,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.6682263343283368,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.778,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6428515876336703,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.7789,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.6134827841251191,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.7603,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.6166557504562269,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.7838,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6581360348610005,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.7735,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5925721410076179,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.7903,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6292226034172892,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.7844,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.7361393543190181,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.8026,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.6207882137657136,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.7843,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.6737289188691014,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.7715,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6439427602847528,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.771,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.6976796686243985,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.8166,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6279320847867842,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.773,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.6500573176229164,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.7897,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.663167597331515,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.7955,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.723174166219946,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.7935,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.701391066772191,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.7847,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.6273877875184655,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.7655,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.6371021859656234,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.7852,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6624679973435736,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.8071,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.6388375904040103,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.7593,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.7013961248759272,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.7821,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6247699438304418,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.7648,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.693135008943615,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.7579,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6325912845686197,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.7829,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.7167442681887659,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.7723,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.6181847393221037,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.7858,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6857968521764412,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.772,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6520810768889314,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.8299,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.7010229389870714,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.7801,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.58920627187464,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.8009,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6583414696958191,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.7651,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.5890705787964754,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.7592,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6019072836773962,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.7745,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.610445952534581,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.7921,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7251948447933639,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.7771,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6411375309034355,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.7648,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6885263988094582,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.7752,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.8034773202790845,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.7965,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.6373643204020462,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.7828,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.8275047771852198,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.7678,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.6390361871343938,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.7823,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6268506796059192,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.7867,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.6106561995234409,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.7918,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6630556075575792,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.7737,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.7084135851007166,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7584,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8419172902879922,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5882,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.6756557826872647,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5995,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.8032275977290458,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.589,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.8856628765284177,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5734,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 16.25569744546536,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5933,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.9608567345215232,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.5956,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.7801636748881827,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5511,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.7967338664614281,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5723,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.7733134209111956,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5708,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7213097092228609,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.6042,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.8201801830146507,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5737,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.6983741030103113,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.6072,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.8358442314636475,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5929,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.8147697310054178,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.5956,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.6958269935176513,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.5569,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.7980219608020527,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5717,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.6807921376234944,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5865,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.7995657777296072,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5721,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.7028608855919712,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5911,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.8432130108080602,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5796,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.6771260891162747,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.5803,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.6788546960094988,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5609,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.6810139098250974,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5775,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.684371865644887,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.5639,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6776759477165379,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.561,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.6500642119889397,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5725,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7065823350158066,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5669,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.6603595598940009,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5688,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6911512348218721,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.548,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.6827838198181937,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5409,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.7404463333178939,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5964,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.7882588465559195,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5749,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6705981766339744,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5473,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.7349903150695317,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.5624,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.7121263811499022,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5831,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.6762787679160464,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5648,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.7153889165605747,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5496,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.6745978515891962,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5888,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.693872692027409,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5782,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.6495590396626207,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.6079,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.6131106572298374,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5826,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.682061268502007,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5906,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.6547869383216933,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5607,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.6603593732886758,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5515,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.635890516075527,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5543,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.6564770052966824,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5678,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.6638841108452911,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5678,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.7434179018221271,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5767,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.648500500198305,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5373,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.6789046273147407,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5837,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.656552327463845,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5388,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.7210336985214922,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5782,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.8205482528235912,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.6,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.6740435150630778,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5936,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8282626858170312,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5766,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7677976213622949,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5906,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.8292430578737281,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5675,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.7871618890818178,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.5571,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.7423996617251396,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5645,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.7245360283771021,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5644,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.703953366128401,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5572,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6969265969224924,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5801,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.6673527544889463,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.582,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.673942509643413,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5444,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.6238482888586189,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5718,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.7938578123209763,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5791,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.6082932581519936,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.5836,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.762530007364864,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5664,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6403901922873099,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5709,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7751178596191762,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.6069,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.772755118322993,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5556,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.7444402297179766,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.5532,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.6988191050629804,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5479,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.7640075254208708,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.598,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.6266578934580058,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5595,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.6487882771999745,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5457,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.6321391674153284,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5763,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.6385592926729851,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5796,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.7164814765661581,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5794,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6399266477177149,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.6006,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.5810513278061431,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5626,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.6357150792487354,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5493,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.7201810969269606,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5664,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5935037694355841,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5909,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.8193630543947713,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.577,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.640702801789411,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5643,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.7327641095087732,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.5704,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.7399463938225568,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5646,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.6371803106330829,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.585,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.7307150648398398,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5896,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.6197803049398256,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5821,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.642148348121148,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.566,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.7410028846188107,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.5579,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.7310127529922015,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5936,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.6606293516263833,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.566,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.6688551066487638,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5984,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.7015736587482053,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.578,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.7185275290030583,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5748,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.6585437237438426,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5861,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.6984985199215237,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5597,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6610405888390912,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.6029,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.7190321399113165,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5682,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.6291246666268816,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5876,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.6845943940353775,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5949,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.6277322903562828,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5766,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.6796309292241134,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5872,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.6229692543372577,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5828,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6618585915260684,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5682,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.646415670543508,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5508,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.6822924946268085,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5877,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.6680525143035462,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.5708,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.5965020466438962,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5643,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6625653874937627,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5926,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.5591697691879005,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5749,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.6545623986532002,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5859,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.5675378595053463,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.5609,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6725272194284734,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5793,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.5955933640055591,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.577,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.6351473187932272,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5668,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.5987736076128526,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5899,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6171923573685534,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5632,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.5764337549405427,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5704,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.5690049082293621,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.5891,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.6363047106318739,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5855,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.5938907251168623,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.562,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.6019179701660279,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5888,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6493520550628022,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.588,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6050247867402664,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5645,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.6067217843486477,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5576,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6439539571978579,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.6112,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.6282625596881151,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5679,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.6723319456744141,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.571,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.603468398720593,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5816,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.7512242370037576,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5621,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6177835924372039,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5677,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6321745884287155,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5825,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.630053838317898,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5626,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.5615998017737761,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5897,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6867048663167408,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5632,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.5467549785664898,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.5538,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.7361260262179837,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5871,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.590885960921104,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.5967,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.6680894021266897,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5836,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6970571690854985,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5469,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5515670744421273,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.585,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.6115110789563288,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.572,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.6426588476884881,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.5762,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6017586112852326,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5741,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.5990295490118738,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5934,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6546450806679445,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5841,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.6478976799829497,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.5747,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6553163521742912,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5727,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.6380468874069141,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5705,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.7019651677885583,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.5917,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.6382252553339951,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5811,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.6788168598868786,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5868,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6041661780540803,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.6205,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6360934101142552,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.6013,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6118405994088386,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5802,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.6447821386276301,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.5906,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.5398216031554207,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5801,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6468124883855235,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5835,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.5761603326307424,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5968,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6349406233872178,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5986,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.5701503932809902,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5957,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5803010115654934,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5739,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.6032451826131848,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.583,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.6125743481251621,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5929,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5433012844620547,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.5863,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6434710115693935,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.5845,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5784904259382722,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.59,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.6220869212472164,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5837,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.7351239562122205,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5719,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.5714171948129009,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5832,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6938142456747738,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5671,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.7849479790951706,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5836,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.5924882755224378,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5578,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6932168271395155,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.6019,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.5710017673461568,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.587,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.585167644562812,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5769,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.6423657838153881,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5963,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.5894353973985177,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.577,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.6258240042367129,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5716,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.7253180047344773,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.6154,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.6654860565779247,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5821,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.7619060531971805,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.5759,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.6083353195043988,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5608,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.6617770505421257,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.5645,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6756517330485747,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.58,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.5938620231100447,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.583,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.5885431516942772,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5946,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.6078802617370986,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5622,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.634678002327514,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5686,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.5839955989635177,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5934,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5561823874760232,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5811,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6493313315364313,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5517,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5958684715242062,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5696,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.6806672543878068,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.589,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.6206583021317464,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5824,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6044852171904841,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.598,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6568950595815862,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.592,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6129820412828476,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5844,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5673638382530058,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5744,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.5818723534832091,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5634,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6531901460998211,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.6018,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5950285259321304,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5571,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.5802091522222553,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.6177,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.585892293204206,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.574,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.5726726500207213,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.5776,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.5534238872932694,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5809,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6151812669363012,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5725,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6045902335163161,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5762,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.5870572554740978,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.5831,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.5601250150188936,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5719,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.598442743984812,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5658,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.5583715336738564,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5925,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.5871190953189237,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5925,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.5833698946791402,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5979,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5598874283311683,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5997,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5401364552306646,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5697,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5949381412862709,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5766,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5427955776923349,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5734,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6214114717898375,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5605,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.5549459528134449,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5613,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.5894144885382049,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.5895,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5429656751223404,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5726,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5975008138073062,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5831,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.5727335641624558,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.593,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5950270891196439,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5716,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5847970179943419,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5687,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.5827011868875309,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5738,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.5597393326859388,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5713,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.5546991129596077,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5738,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5843969347051255,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5614,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.598576916021465,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5596,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.6349223343941945,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.584,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.6570996483268494,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.5736,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5339158407221707,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5588,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.639493884997266,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5824,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.6084059676220311,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5818,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6487168390105794,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5515,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6582485463391419,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.6111,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5486856788282561,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.567,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.7396236893974113,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5897,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5477609336921122,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5739,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.6443882613609895,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5725,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.6037574644742653,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.5088,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.70577759315347,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3714,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.6412190394700527,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3466,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 1.0304463469613097,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3312,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.5896705705585356,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3256,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.75597316822809,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3603,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.6552338371033397,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.352,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.6078989771899074,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3319,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6312418582972059,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3519,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.6237342274593018,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.35,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.6144412732432626,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3492,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.5959234145236924,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.367,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.5610819127464566,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3511,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.5148599447206842,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3712,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.5856520985621919,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3501,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.5473938709420568,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.354,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.5864676357355998,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3602,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.5734627262053782,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3559,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.5407357875242259,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3357,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.5469857515591342,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3556,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.5441605116415775,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3365,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.5298588867110069,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3445,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.5222993296283089,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3431,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.5553278439400993,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3523,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.5293966713963861,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3632,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.5565326871933542,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3724,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.5082161216928157,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3363,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5599515344896282,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3196,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.5550101444042153,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3383,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5851384865007954,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3505,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.5340275874499217,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3147,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.6028327839997195,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3625,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.5577175776653744,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3634,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5413398767636938,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3365,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5489932849124322,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3126,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.545858946288647,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3204,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5581408178714279,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3541,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.5522958703016551,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3608,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.5552394441925566,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3768,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.565590720139514,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3469,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.5404717236182482,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3588,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.5925655216302819,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3336,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5284868579895091,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3476,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5985217245243283,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3366,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.5521436963625075,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3568,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.5345171203744797,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3438,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.5146467509464381,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3539,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.5315449635268847,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.365,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.5318970961449797,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3773,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5234175316294611,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.324,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.5441723175550004,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3375,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.48858200895164283,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3474,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.5599791057394341,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3722,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.5434689186560627,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3491,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5616427608925829,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3093,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5432139182013139,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3511,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5083760296610442,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3589,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.5431127945341345,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3819,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.5169105802350284,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3684,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5314187615801296,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3474,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5170976765176111,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3421,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.5183443456977562,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.354,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.5051108145132757,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.358,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.500335328519008,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3777,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.5482023512762354,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3368,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5401421200587564,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3409,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5354256631315133,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3628,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5335678362797601,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3577,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5336203682436158,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3511,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5478031005694795,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3711,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.5183787786149585,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3438,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5160382089614298,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3395,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5376962726773228,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.351,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5144107925553082,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3601,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5277087890350307,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3599,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.547716839985386,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3785,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5228646346668387,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3546,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.5218146909085264,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3569,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5156652542357953,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3578,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.499567114140454,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3635,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5448291636482103,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3595,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5222325999508463,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3746,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.5149759930504018,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3417,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.4925706917904922,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3416,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.5431117348372243,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.374,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.5145349537123449,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3486,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5009725846098005,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3445,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.5419871343889099,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3621,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5097837101046062,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3546,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.5177362950246309,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3345,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.5319874859046423,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3682,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.5501604639255262,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3397,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.539740987130656,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3332,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5433165501365943,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.364,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4926478119985206,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3322,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.5369356868087957,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3689,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5281477077732376,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3164,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5512754816470092,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3672,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.5236932698733044,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3378,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5667882266051862,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3638,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5741442117490798,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3694,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.535022109616254,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3432,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5827955715530141,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3693,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5175951637136953,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3448,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5551327423117771,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3483,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5243051323326269,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3737,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4939582158121821,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3322,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.53165160058163,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3513,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5348440361088205,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.335,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5059224268136856,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3656,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5855814987787656,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.359,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5541093171445991,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3724,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.536901100065429,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3708,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5477249776808333,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3582,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.5602072206373201,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3433,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5333937917572233,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3563,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5281947326484094,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3512,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.5289289062530395,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3411,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5256719729034605,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3419,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5424531726108885,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3726,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5058959130306026,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3442,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5135684733676548,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3748,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.533545405034479,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3794,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.507983771678424,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3678,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.49696277714575504,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3476,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.5372021216310732,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3385,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5345239742684599,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3369,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5226152992962165,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3681,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5379660257968912,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3561,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5029217894588218,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3488,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5339971866186536,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3627,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.512229726154632,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.351,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.48512369551423257,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3438,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.4913701852537671,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3515,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5008594972847723,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3256,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5479078999485384,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3492,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.509448384958121,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3553,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5423269754810375,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3421,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.528496034731418,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3498,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5120673731310237,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3545,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5272959684353788,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3562,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.518336637042819,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3249,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.5325198341637791,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3383,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5295143664813663,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3481,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5052344882593199,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3297,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5423628148064337,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3418,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.5102195711959673,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3563,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.5328911298161293,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3423,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5221336601426599,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3543,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5127088239599747,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3626,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5122738972839401,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3505,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.49416156949727896,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3655,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.5197354223170099,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3447,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.5044394541156211,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3546,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5468117232312578,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3478,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.5241073962572423,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3743,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.48872984354405524,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3422,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.5195763178586233,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3433,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.491824873178817,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3342,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.495784181024009,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3413,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.521965819662273,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3582,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5121307051291832,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3463,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.5112639639375739,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.37,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5122656776215798,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3484,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5362526375379877,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3349,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.523186886631131,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3405,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5299190667293312,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3361,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5095024700472262,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3668,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.507466547386342,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.359,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5429219295628898,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3399,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5343157694881103,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3331,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.5103406460325961,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3632,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.525086321559713,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.372,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.4974836002641573,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.3734,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5394704116240879,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.356,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.5251879518662342,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3438,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5139508006440136,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3653,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.5545679605839011,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3459,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.5051759190660727,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3415,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.49876326471770654,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.343,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5127663770414775,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3485,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5388402240706162,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.343,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5058451313480682,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3489,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5432661575833849,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3518,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5293369695234669,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3647,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.49771546772704806,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3479,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5299738278004013,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3556,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.48597615192522003,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3552,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5270458154444475,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3329,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5126066636393667,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3386,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5054250460463976,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3335,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5651050208017063,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3467,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.522845171299228,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3637,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.5242525620612795,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3448,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.5613864354533996,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.344,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5232146088149647,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3867,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.500312431197609,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3282,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5070629687074536,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3592,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5033718585853024,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3359,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.4985733332892675,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3548,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.5195595094149409,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3525,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.5116342290676039,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3302,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.5185733343277323,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3469,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.4835506564760782,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3408,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.566934179135919,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3278,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.4935962558919029,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.342,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5078135410154994,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3569,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.4981959495920278,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3486,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.5214657563057749,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3292,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5005953633298853,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3219,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.5363176582212271,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3808,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5078216012678386,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3702,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.4964111013699246,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3333,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5190582839310105,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3434,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.5254602834974944,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3478,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5067326009256302,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.335,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5121653995511586,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3404,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.522605298570992,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3585,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.502612209948515,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3554,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.4949230391776719,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3494,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5338271439856402,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.354,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.5157455341050571,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3344,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5411187371537644,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.384,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5150857575410472,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3642,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.5411277617634276,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3555,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5139717505614886,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3433,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.48948965153268104,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3729,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.5429674744003463,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3464,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5076426981500822,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3451,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.5279805164363998,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3588,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.548030475249155,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3477,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.4933217655558008,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3562,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5264602288962826,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3574,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5474740904551181,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3453,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.4980632531041245,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3638,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.47522494653940195,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3442,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.495303080735989,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3382,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5084553502989796,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3485,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.4873105586568391,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3479,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.4904935882766413,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.327,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.5093652392518528,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3584,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.477209357080153,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3434,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.507196768830595,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3466,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.49705606238230504,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3381,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.4882709693902527,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3293,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.5043668235293385,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3442,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.49668416706997887,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3408,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.4850284990191855,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.3197,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.5717637829966975,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2449,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.43574032964156284,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1996,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.4582527632035172,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1872,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.6738735515409583,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2146,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.6611661233054077,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2102,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.46322329122005124,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1982,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.4876269696818547,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.195,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.538988084262464,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2176,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.48857727959144515,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1657,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.42737417051826393,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.204,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.4385910880340555,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2219,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.47224496667063043,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2194,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.5070003159712432,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.194,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.43554727651849084,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2001,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4172519973627278,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2155,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.40798726943924374,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1901,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.4417445687099887,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.1827,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.4200904665927483,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1886,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.4077348606624145,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2043,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.4150860355430705,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2181,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.42869498309793314,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1993,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.41417899655555523,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.206,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.40404527316509864,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1828,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.4061248505893929,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2006,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.3981533716857377,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.215,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.42340128227757434,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1966,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.4023357242304229,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1732,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.4088791326573225,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2008,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.4439779720114971,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2187,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.41687246131317196,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2194,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.40900618450330617,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1952,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.4225274361885642,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2288,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.3998954112708777,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1901,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4008883340542458,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1956,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.4127661790518433,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1852,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.411570708820223,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1736,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.41910854376750484,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2136,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.40370832743177226,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2028,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.3983471977628199,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2131,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.41831417946131494,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2032,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.39623256016255576,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1963,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.39116526701737214,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1983,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4184944346104359,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.1871,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.41628001821116767,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1975,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.4088669906090832,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.1995,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.40587379235627347,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2118,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.4113869874977336,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1829,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.40541841152190633,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1766,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.39017033912455945,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2245,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.4178826587379696,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.177,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.3998185357283628,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1832,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.413560242657442,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2008,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.40113802223898054,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1937,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4044168406319632,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1894,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.40134080113085674,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2246,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.4277944133240199,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1732,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.38902750006792525,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2099,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.3968945748576292,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.196,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.4104958563193352,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.184,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3954659586598155,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.216,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.4003595643529471,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1889,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3883145860602173,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1896,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.4150349541880968,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.182,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.39049366323114915,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1883,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.3916466721031382,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2004,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.41885144115745016,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1948,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.4050105478426393,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.199,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.3975117785769752,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1859,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.38011619355032916,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2045,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.40561501904687874,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1849,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.4055613603585829,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1858,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.4035690945622511,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1746,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.39966056004099965,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2211,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.402088676567704,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2085,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.39091914653164805,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2016,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.3875304037338862,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1849,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.3940204548853163,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1998,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.3809319968381036,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2012,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.38454983696912703,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2029,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.4081880024674592,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1901,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.3906340986526364,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1889,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.3825865994828239,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1834,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.3964295472077241,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1952,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.3972194454632369,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.208,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.4045263128069779,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1782,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.3892142646769035,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2087,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.40695117047924745,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.2044,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.38665222531797727,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2029,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.3948986001987699,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2084,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.3841374809949284,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1948,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.4216316743998641,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1916,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.3951451820005121,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2144,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.4032080820849941,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1688,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.41129196517002536,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2064,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.38089714660044915,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1718,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.39131124636140624,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2018,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.39691445504161976,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1978,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.39665591547937723,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1843,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.39308291092588327,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1824,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.4010323312105742,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1931,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.3891944823944482,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1926,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.39335003937775603,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2022,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.38880910988367823,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1928,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.38269064297737676,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.1863,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.39206804668480943,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.189,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.39142111826068027,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1849,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.39414551283085136,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.214,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4266527569164646,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1875,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.37909692669226835,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2055,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.392657516059937,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1825,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.38062120572729996,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1927,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.40774078265481883,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1706,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.3708584204432429,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1824,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.38591978030239066,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1804,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.3872761017975701,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.1789,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.3717460117395034,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2051,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.3979394626361227,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1911,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.3835948400227757,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1944,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.3951098181758579,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2029,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.39518499646753585,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2012,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.3852151907320285,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1995,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.4105364534725083,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1827,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.4006677118336395,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.1819,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.4126994596745534,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1946,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.4081372449868042,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1963,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.37197051808466375,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1915,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.39049435793482246,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1866,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4105155561378278,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2022,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.3922046644698891,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1956,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.36776331560699926,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.181,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3881034642304604,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1997,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.4008416831230104,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.195,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.38785883617854144,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2179,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.3942134397939783,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.1977,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.39243576761346805,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1813,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.3772494620116382,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1874,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.3913552632896342,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1794,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.4078990137075546,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1897,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.39786121822482723,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1626,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.3950576605747718,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1857,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.40355817311220427,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1924,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.38887175092762916,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.209,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.37433929940388805,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1922,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.38412243810223373,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1976,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.38935720928753154,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1648,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.38599591219364765,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1929,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.37909410288021717,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2166,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.4023074433563224,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1728,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.3892348418960725,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2005,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.3893700736495359,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1665,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.39460911992694436,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1956,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.3964781299417055,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2043,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.39028177022238536,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1783,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.38183209782517974,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.1628,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.38392222869735143,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.1799,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.3875387420973212,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1785,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.3904726320978955,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2082,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.39826561311673025,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2029,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.388402602486022,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1972,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.396003440081996,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1695,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.38759476110162205,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1785,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4028848607537817,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1861,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.3943761938728602,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1997,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.3860438096487179,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1855,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.38308129310051287,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2005,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.3758569363754197,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.194,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.390763543360762,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1964,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.3730276100437758,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1766,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.41056049795107274,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1965,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.39179469935487915,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1902,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.38764899298298067,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2001,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.4034820134692344,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1736,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.3973262950955315,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2052,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.39437752241743607,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1982,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.39548931126073594,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2097,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.39535480316364513,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1952,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.3831252689871781,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1915,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.36977264246128777,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.185,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.3914711088268367,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.208,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.3952081997666062,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2059,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.3949706855099044,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2085,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.36858290293887763,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2151,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.40894259881679607,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2252,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3865112381975527,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1967,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.3854221476854259,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1763,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.3829632969944751,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2067,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.38558927887405664,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1931,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.3925253400405845,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1615,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.3902963845535035,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1758,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.3935234363763406,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1901,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.37891294201696435,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1888,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.3825355757711077,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.172,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.3768703207878216,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1938,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.41785036203351145,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.183,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.39380750053132235,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1806,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.386055942197102,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1628,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.39314229182132915,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2092,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.3857959020759943,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1974,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.380766874476722,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1719,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.3803956492393655,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1872,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.3836881837956418,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1879,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3887138237302928,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2005,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.3748464890177696,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1796,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.39012598252139785,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1723,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3972487743308296,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1989,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.38777693265098717,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1766,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.38169458397770284,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1881,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.37894646183376496,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1673,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.3718368262837119,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1737,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.39527618896186867,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1868,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.3755004366738763,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1892,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.3918468174668289,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1835,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.40096601234573204,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1737,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.38610759629031177,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.186,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.37923740774014764,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1808,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.39067511423952195,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1516,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.3743934259395775,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1733,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.3741096633859345,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1694,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.3735055641855606,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.172,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.3757210240652418,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.1807,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.37200607803093494,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1531,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.3980853412501174,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1677,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.3697589418193144,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1939,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.3916732011682174,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1642,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.3658794059171595,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1776,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3731941402920819,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1942,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.3763523395612373,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2194,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.3801833801137943,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.1817,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.37226031850215396,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2147,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.3869345678398528,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1752,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.3810530118327359,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1745,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.3877194684611994,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1925,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.3708553454864818,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1651,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.3851007815721319,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2041,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.3810057601988319,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.1759,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.3891670808927755,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1858,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.3938399956399788,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1697,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.3918131940706886,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1856,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.3838121813898066,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1802,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.3902372576026202,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2176,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.38267037243980234,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.1913,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.37666609123798256,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1768,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.36852702838280643,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1823,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.38228314681232234,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1848,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.37826302906553866,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1893,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.3711951755059744,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2103,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.3592145347595648,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.146,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.4216879977975434,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.129,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.34798288943308453,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1216,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.3195694827989809,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.1067,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.27599896185472594,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.0991,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.2818117727723549,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1384,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.3159241129044614,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1201,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.36559246673017787,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1271,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.3827454727986112,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1141,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.3979587837466413,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.145,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.36076446189267586,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1176,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.33189317036452753,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1197,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.28834355658867217,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1092,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.30552502324340947,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1328,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.3290644789747052,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1246,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.3250236221866049,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1223,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.33125148153832257,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1441,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.29983557858689636,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.1391,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.2665264504834675,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1205,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.2684956036778013,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1346,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.2707552610927497,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1342,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.29762640593315237,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1274,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.29147339496391333,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1231,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.295430686082123,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1374,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.31025290065520306,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1322,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.26842067720643054,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1089,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.26007310145396684,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.1257,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.27696256172130346,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1285,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.27337397015815584,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1196,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.2712993249631705,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.154,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.2844331884293529,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.113,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2879831540812325,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1515,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.29373311232543375,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1135,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.2796807502952066,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1291,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.2768471537617241,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1447,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.28384709488858156,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.105,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.2718039068350403,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1269,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.2879700675128376,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1215,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.28422718614042153,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1236,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.28347076531347104,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1171,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.3036528633568336,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1481,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.27655488604862977,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1243,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.2730215828637935,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1514,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.2711668349290487,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1139,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.2725700511654941,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1284,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2740404375640797,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1428,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.2859777279590289,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1407,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.25980376732836613,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1235,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2719300746589577,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1141,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.26498116036891395,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1149,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.258128793511937,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1146,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.2726776423966777,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1152,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.27069890672444186,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.135,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.26237183883996845,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.1288,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.2671610554483958,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1478,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.26184276982119514,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1243,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.2590584366438277,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1048,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2795034050755459,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1136,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.2612721847590388,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1265,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2714600301048136,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1126,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.27361557939050996,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0928,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.28504888424746155,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.127,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.27844744308939545,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1143,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.2736363096414075,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1347,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.28163041729622906,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1193,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.272903759119448,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1189,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.2616225490299821,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1115,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.26731703469256424,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1138,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.2598306783147096,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1276,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.2768303910082332,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.1215,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.2699325317004273,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1252,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.2570338365430844,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1102,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.2692739677285256,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1192,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.2822073863867715,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.1416,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.27161039849229285,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.1336,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.2547022217729296,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1307,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.2721232945969668,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1052,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.27533726619569343,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1148,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.2840537375760081,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1194,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.28222619438527724,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1373,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.24687655355504834,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1076,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2746164306495859,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1221,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.2714348469665864,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1204,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.2731281966915283,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1185,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.26159005722870865,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.103,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.2640011710736412,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.119,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.28096544110529753,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.177,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.27502157938471256,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.114,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.2701911989985527,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1273,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.28605894224031875,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1361,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.2867566768032954,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1209,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.26863270114052784,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1282,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.2750163059080665,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1177,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.2823874093978708,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1208,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.26339774073750594,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1538,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.26772426748874506,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1202,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.265328964132435,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1395,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.27312993788581097,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.1173,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.28404327713479915,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1399,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.2737614632593357,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1215,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.26654166428957476,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1285,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.2735855246154714,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1384,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.26086453569548124,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1294,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.26817086939966356,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1271,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.2766578929079356,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1252,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.2630616333701599,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1233,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.295576727624469,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1268,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.25602798189121406,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1049,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.27431019437129905,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1459,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.2587536897707474,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.129,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.27137127918895826,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1208,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.26781626238883777,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1082,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.2745009722293633,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1411,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.2713366135430676,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0958,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.2619688606764706,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1142,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.2808475778457547,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1284,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.2836072328404829,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1253,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.2740027770906232,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.1121,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.2660949353202542,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1141,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.288960033611544,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1288,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.2771529555258204,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1456,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.26906625122996697,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1245,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.2571385770864678,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1236,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.26586885939729354,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1093,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.25980229949694245,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.0994,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.2746710692065458,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1235,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.2758630690664129,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1207,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.26856726174018763,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1041,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.2743227253003253,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1458,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.26946402252855434,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1223,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.28036599864533246,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.145,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.27404123517465123,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1218,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.26139510509956043,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1077,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.27251768223379796,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1313,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.2807438919775813,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1504,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.2633570342557104,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1287,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.2858427011293434,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1411,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.26597114109841635,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1031,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.27089084056305157,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.145,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2769038033523164,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1353,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.27793655923220245,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1188,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.27432595998800774,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1263,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.2710682289325277,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1071,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.27956109258681466,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1204,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.26571867801598414,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1136,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.26231800555146156,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1547,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.2617702875728245,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1155,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.2717990909908089,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1316,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.2864577471604053,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1274,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.2731865728994914,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1017,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.26933868083514567,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1256,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.27152014836799354,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1059,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.25823570434749626,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1161,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.25298146417485723,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1035,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.26730290165913284,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.11,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.2918246979359156,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1256,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.2758374558274925,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1189,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.26180753045590727,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1309,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.26988903511736084,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1041,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.26482107479050043,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1233,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.2578248950432625,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1062,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.26031346608045686,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1198,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.2606977771548412,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1132,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.2702598994885059,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1023,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.275001019182037,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1224,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.28304320122760135,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1242,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.26365952743428817,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1193,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.2610830200736935,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1092,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.2599891874054502,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1533,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.27279652780510594,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1171,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.28792782713696863,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.137,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.2740154056675449,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1209,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.27021650636780264,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1344,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.25633117209717227,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1269,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.2726734082203001,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.126,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.27034284906547174,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.106,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.2668268291159694,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1306,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.26883413871255046,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.0901,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.2647147430674471,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1069,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.2676081060103831,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1038,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.29522400846371144,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1307,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.2620076768906162,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1326,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.2626758412432568,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1042,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.28044241787247576,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1304,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.27710680277510696,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1282,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.2802988316575978,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1404,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.2598149830709728,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0959,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2758618413265175,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1197,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.26575086035131734,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1348,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.28524594219880084,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1322,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.26958492199202216,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1252,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.27102934307778453,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1283,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.27524451522935717,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.117,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.2730204325264804,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1283,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.28341957632333364,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1271,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.2942054653571622,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.1359,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.26844289507754465,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1316,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.27440745269289557,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1249,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.2734125840671842,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.0961,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.27884489781556127,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1289,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.26462855386463535,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1077,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.2836568916819393,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.132,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.27999496865862233,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1245,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.27107599590514986,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1198,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2773989267589204,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1365,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.28366073467357306,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0992,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.3049961718896443,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1141,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.27020891927319585,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1154,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.28489326166803264,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1319,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.27296357789820164,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1186,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.27161231159145094,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.129,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.2759621258683222,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1266,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2796987462434727,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1474,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.27705863650746254,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1049,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.2629419636701384,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1012,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.2727974130533818,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1298,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.26012511247520964,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0946,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.28950119330739627,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1216,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.27136054592389297,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1395,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.27328592213318986,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1177,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.2510394367682423,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1107,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.2717315002730077,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1095,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.2687837651642364,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.1071,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.2589057908817606,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1168,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.2611106861793302,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1226,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.27243251970772087,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1115,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.2806767987540215,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1383,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.27291659591816625,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1214,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.25866286878938,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1159,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.274467540660026,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1195,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.27777803949188284,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1372,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.26918318441068195,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1145,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.25823768979496486,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1187,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.27805047568617564,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1273,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.26168514869646564,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1424,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.2646028290560397,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1262,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.2755648389783531,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1198,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.2795091688797782,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1271,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.2809568637801747,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1349,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.2803909701655631,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1252,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.2732193153263027,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.122,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.25822888724782295,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1138,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.2698755889030994,
+      "learning_rate": 0.0,
+      "loss": 0.1123,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 1.3052586975687885e+18,
+      "train_loss": 0.4145879797455741,
+      "train_runtime": 24797.6648,
+      "train_samples_per_second": 6.372,
+      "train_steps_per_second": 0.05
+    }
+  ],
+  "logging_steps": 1,
+  "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": 1.3052586975687885e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}