diff --git "a/checkpoint-20924/trainer_state.json" "b/checkpoint-20924/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-20924/trainer_state.json"
@@ -0,0 +1,146497 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999681396756619,
+  "eval_steps": 500,
+  "global_step": 20924,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.214809069965307,
+      "learning_rate": 2.390057361376673e-08,
+      "loss": 1.6348,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.59734848256109,
+      "learning_rate": 4.780114722753346e-08,
+      "loss": 2.5645,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.893297283030142,
+      "learning_rate": 7.17017208413002e-08,
+      "loss": 1.6387,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.592278396272854,
+      "learning_rate": 9.560229445506692e-08,
+      "loss": 0.9863,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.291241122952738,
+      "learning_rate": 1.1950286806883366e-07,
+      "loss": 2.6953,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.67999961663868,
+      "learning_rate": 1.434034416826004e-07,
+      "loss": 2.8906,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9120736517090675,
+      "learning_rate": 1.6730401529636714e-07,
+      "loss": 1.7441,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6720238188758527,
+      "learning_rate": 1.9120458891013384e-07,
+      "loss": 2.3086,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.830784854927664,
+      "learning_rate": 2.1510516252390057e-07,
+      "loss": 3.3496,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.268575614746341,
+      "learning_rate": 2.390057361376673e-07,
+      "loss": 2.207,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.624460839267948,
+      "learning_rate": 2.629063097514341e-07,
+      "loss": 1.8984,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.21154328136634,
+      "learning_rate": 2.868068833652008e-07,
+      "loss": 2.6621,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.233767949293284,
+      "learning_rate": 3.1070745697896753e-07,
+      "loss": 1.3516,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.307043410515324,
+      "learning_rate": 3.346080305927343e-07,
+      "loss": 1.707,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.401300997692752,
+      "learning_rate": 3.5850860420650093e-07,
+      "loss": 3.1289,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.078238235332442,
+      "learning_rate": 3.824091778202677e-07,
+      "loss": 3.3867,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.906148192032673,
+      "learning_rate": 4.063097514340345e-07,
+      "loss": 2.1758,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.204491829120101,
+      "learning_rate": 4.3021032504780114e-07,
+      "loss": 1.8164,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1440172995656788,
+      "learning_rate": 4.5411089866156795e-07,
+      "loss": 1.6777,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9784610330550314,
+      "learning_rate": 4.780114722753346e-07,
+      "loss": 1.8867,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.7569182157985,
+      "learning_rate": 5.019120458891013e-07,
+      "loss": 1.2598,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.561450793280567,
+      "learning_rate": 5.258126195028682e-07,
+      "loss": 3.125,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5324734619180473,
+      "learning_rate": 5.497131931166347e-07,
+      "loss": 1.8086,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.250792934273364,
+      "learning_rate": 5.736137667304016e-07,
+      "loss": 1.5645,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.092882836203435,
+      "learning_rate": 5.975143403441683e-07,
+      "loss": 1.6328,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.954202202708203,
+      "learning_rate": 6.214149139579351e-07,
+      "loss": 3.4805,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.8641243935288765,
+      "learning_rate": 6.453154875717018e-07,
+      "loss": 1.4443,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.753503272961867,
+      "learning_rate": 6.692160611854686e-07,
+      "loss": 2.9395,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1339944716881325,
+      "learning_rate": 6.931166347992353e-07,
+      "loss": 1.9883,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.398358720784021,
+      "learning_rate": 7.170172084130019e-07,
+      "loss": 1.6797,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.661583603255156,
+      "learning_rate": 7.409177820267687e-07,
+      "loss": 2.5488,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4755294040943827,
+      "learning_rate": 7.648183556405354e-07,
+      "loss": 1.7656,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5805979572485844,
+      "learning_rate": 7.887189292543021e-07,
+      "loss": 1.7539,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.571604779647267,
+      "learning_rate": 8.12619502868069e-07,
+      "loss": 2.3848,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.468180724301122,
+      "learning_rate": 8.365200764818357e-07,
+      "loss": 1.1797,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.621501504592167,
+      "learning_rate": 8.604206500956023e-07,
+      "loss": 2.3945,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3827332802321894,
+      "learning_rate": 8.84321223709369e-07,
+      "loss": 1.3809,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.257986923898525,
+      "learning_rate": 9.082217973231359e-07,
+      "loss": 1.5059,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.715798764469675,
+      "learning_rate": 9.321223709369026e-07,
+      "loss": 3.8711,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.2434768469791,
+      "learning_rate": 9.560229445506693e-07,
+      "loss": 2.1094,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.6656513800535535,
+      "learning_rate": 9.799235181644359e-07,
+      "loss": 1.6484,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.553901736427874,
+      "learning_rate": 1.0038240917782027e-06,
+      "loss": 1.7578,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.7425148045734815,
+      "learning_rate": 1.0277246653919695e-06,
+      "loss": 1.3516,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.035229291344407,
+      "learning_rate": 1.0516252390057363e-06,
+      "loss": 1.8438,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.75391976776777,
+      "learning_rate": 1.075525812619503e-06,
+      "loss": 2.8652,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.666314337336278,
+      "learning_rate": 1.0994263862332695e-06,
+      "loss": 2.8652,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.189780411971908,
+      "learning_rate": 1.1233269598470365e-06,
+      "loss": 1.4512,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.839000092697194,
+      "learning_rate": 1.1472275334608031e-06,
+      "loss": 2.2812,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 39.07231751628959,
+      "learning_rate": 1.17112810707457e-06,
+      "loss": 3.4238,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.597237464162532,
+      "learning_rate": 1.1950286806883365e-06,
+      "loss": 1.7793,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.35116158034682,
+      "learning_rate": 1.2189292543021033e-06,
+      "loss": 1.8574,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.709503829269362,
+      "learning_rate": 1.2428298279158701e-06,
+      "loss": 2.7852,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.928650643133537,
+      "learning_rate": 1.2667304015296367e-06,
+      "loss": 1.6953,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.763053237931012,
+      "learning_rate": 1.2906309751434035e-06,
+      "loss": 2.5488,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.904061501468861,
+      "learning_rate": 1.3145315487571701e-06,
+      "loss": 2.3887,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.290611989091765,
+      "learning_rate": 1.3384321223709371e-06,
+      "loss": 2.4062,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.019967454295645,
+      "learning_rate": 1.3623326959847037e-06,
+      "loss": 2.0215,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1068118471868877,
+      "learning_rate": 1.3862332695984705e-06,
+      "loss": 1.5762,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.659630489188641,
+      "learning_rate": 1.4101338432122371e-06,
+      "loss": 1.9824,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3940399410252775,
+      "learning_rate": 1.4340344168260037e-06,
+      "loss": 1.7676,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.567691516105029,
+      "learning_rate": 1.4579349904397707e-06,
+      "loss": 2.7656,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1368503946729067,
+      "learning_rate": 1.4818355640535373e-06,
+      "loss": 1.3555,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.931392320481697,
+      "learning_rate": 1.5057361376673041e-06,
+      "loss": 1.4473,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0847389694401244,
+      "learning_rate": 1.5296367112810707e-06,
+      "loss": 2.0332,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 56.04332909747711,
+      "learning_rate": 1.5535372848948375e-06,
+      "loss": 3.8672,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.967877176061338,
+      "learning_rate": 1.5774378585086041e-06,
+      "loss": 1.5918,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7476469922900693,
+      "learning_rate": 1.601338432122371e-06,
+      "loss": 1.4102,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.316577637743098,
+      "learning_rate": 1.625239005736138e-06,
+      "loss": 1.6816,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.156781559536487,
+      "learning_rate": 1.6491395793499044e-06,
+      "loss": 1.5605,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.63178587646176,
+      "learning_rate": 1.6730401529636714e-06,
+      "loss": 2.3086,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.396286443393308,
+      "learning_rate": 1.6969407265774378e-06,
+      "loss": 1.0771,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2459877868787,
+      "learning_rate": 1.7208413001912046e-06,
+      "loss": 1.6738,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.591446460593539,
+      "learning_rate": 1.7447418738049716e-06,
+      "loss": 2.0273,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.143961396971056,
+      "learning_rate": 1.768642447418738e-06,
+      "loss": 2.1289,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4053455981613614,
+      "learning_rate": 1.792543021032505e-06,
+      "loss": 1.9219,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.166554149254152,
+      "learning_rate": 1.8164435946462718e-06,
+      "loss": 1.6621,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.3929037246658185,
+      "learning_rate": 1.8403441682600384e-06,
+      "loss": 1.9805,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.340505017744892,
+      "learning_rate": 1.8642447418738052e-06,
+      "loss": 2.9082,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.872916730163409,
+      "learning_rate": 1.8881453154875718e-06,
+      "loss": 1.1016,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7214424569198012,
+      "learning_rate": 1.9120458891013386e-06,
+      "loss": 1.5938,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.442991970295973,
+      "learning_rate": 1.9359464627151056e-06,
+      "loss": 1.3418,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.63504216988193,
+      "learning_rate": 1.9598470363288718e-06,
+      "loss": 2.0391,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.482326003826433,
+      "learning_rate": 1.983747609942639e-06,
+      "loss": 1.0898,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.062229111115341,
+      "learning_rate": 2.0076481835564054e-06,
+      "loss": 2.0195,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7186425724948107,
+      "learning_rate": 2.031548757170172e-06,
+      "loss": 1.5547,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.507081922966028,
+      "learning_rate": 2.055449330783939e-06,
+      "loss": 1.5371,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.114836822804633,
+      "learning_rate": 2.0793499043977056e-06,
+      "loss": 2.0723,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.365890642625698,
+      "learning_rate": 2.1032504780114726e-06,
+      "loss": 1.4863,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.963264256279752,
+      "learning_rate": 2.1271510516252388e-06,
+      "loss": 1.7695,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2347006518671773,
+      "learning_rate": 2.151051625239006e-06,
+      "loss": 1.2461,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.320859463991899,
+      "learning_rate": 2.174952198852773e-06,
+      "loss": 2.0469,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.871978664913998,
+      "learning_rate": 2.198852772466539e-06,
+      "loss": 1.5625,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.16473789066935,
+      "learning_rate": 2.222753346080306e-06,
+      "loss": 2.5254,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.382679620661222,
+      "learning_rate": 2.246653919694073e-06,
+      "loss": 1.3203,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.466515692656698,
+      "learning_rate": 2.2705544933078396e-06,
+      "loss": 2.1445,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.29161408593983,
+      "learning_rate": 2.2944550669216062e-06,
+      "loss": 2.5596,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.653309022623574,
+      "learning_rate": 2.318355640535373e-06,
+      "loss": 4.2109,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5574680327117063,
+      "learning_rate": 2.34225621414914e-06,
+      "loss": 1.3672,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.43314024859454,
+      "learning_rate": 2.3661567877629064e-06,
+      "loss": 3.6055,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.38098288422426,
+      "learning_rate": 2.390057361376673e-06,
+      "loss": 2.1152,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.893447854379886,
+      "learning_rate": 2.41395793499044e-06,
+      "loss": 1.5801,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.87792473897553,
+      "learning_rate": 2.4378585086042066e-06,
+      "loss": 2.1152,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4099162957889613,
+      "learning_rate": 2.4617590822179732e-06,
+      "loss": 1.5234,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.704766662086536,
+      "learning_rate": 2.4856596558317402e-06,
+      "loss": 1.3301,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.186928244725244,
+      "learning_rate": 2.509560229445507e-06,
+      "loss": 1.4062,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.307459565712309,
+      "learning_rate": 2.5334608030592734e-06,
+      "loss": 1.832,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9605200480790743,
+      "learning_rate": 2.55736137667304e-06,
+      "loss": 1.5176,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.281236639056064,
+      "learning_rate": 2.581261950286807e-06,
+      "loss": 1.6211,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.634923700663984,
+      "learning_rate": 2.6051625239005736e-06,
+      "loss": 0.9688,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.292970471826218,
+      "learning_rate": 2.6290630975143402e-06,
+      "loss": 2.0742,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.353260878994617,
+      "learning_rate": 2.6529636711281073e-06,
+      "loss": 1.9297,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.780155779193921,
+      "learning_rate": 2.6768642447418743e-06,
+      "loss": 1.9668,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.850038052439117,
+      "learning_rate": 2.7007648183556404e-06,
+      "loss": 1.5879,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.434016976298082,
+      "learning_rate": 2.7246653919694075e-06,
+      "loss": 3.6328,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.419796353118164,
+      "learning_rate": 2.748565965583174e-06,
+      "loss": 1.625,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.230634663182383,
+      "learning_rate": 2.772466539196941e-06,
+      "loss": 1.3887,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.002797114616541,
+      "learning_rate": 2.7963671128107077e-06,
+      "loss": 1.9121,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.312411315901868,
+      "learning_rate": 2.8202676864244743e-06,
+      "loss": 1.7031,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.53994104587085,
+      "learning_rate": 2.8441682600382413e-06,
+      "loss": 1.8672,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.339486643934481,
+      "learning_rate": 2.8680688336520075e-06,
+      "loss": 1.3369,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2207193827491905,
+      "learning_rate": 2.8919694072657745e-06,
+      "loss": 1.2197,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5255326430016383,
+      "learning_rate": 2.9158699808795415e-06,
+      "loss": 1.4463,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.3338042970855435,
+      "learning_rate": 2.939770554493308e-06,
+      "loss": 1.7246,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5155215099785098,
+      "learning_rate": 2.9636711281070747e-06,
+      "loss": 1.8379,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.392079893629979,
+      "learning_rate": 2.9875717017208413e-06,
+      "loss": 1.6387,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.81468313278632,
+      "learning_rate": 3.0114722753346083e-06,
+      "loss": 1.7695,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.130879731752884,
+      "learning_rate": 3.035372848948375e-06,
+      "loss": 2.0215,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.666295028745857,
+      "learning_rate": 3.0592734225621415e-06,
+      "loss": 1.9688,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.303760954163284,
+      "learning_rate": 3.0831739961759085e-06,
+      "loss": 1.9941,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.8465646748959434,
+      "learning_rate": 3.107074569789675e-06,
+      "loss": 1.4785,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.659854983571451,
+      "learning_rate": 3.1309751434034417e-06,
+      "loss": 1.3086,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.955103470861077,
+      "learning_rate": 3.1548757170172083e-06,
+      "loss": 1.8691,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.42494191262361,
+      "learning_rate": 3.1787762906309753e-06,
+      "loss": 1.8887,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.948517879969115,
+      "learning_rate": 3.202676864244742e-06,
+      "loss": 2.1758,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.049235358876679,
+      "learning_rate": 3.2265774378585085e-06,
+      "loss": 1.9512,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.749627939380344,
+      "learning_rate": 3.250478011472276e-06,
+      "loss": 1.4258,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1343138588235515,
+      "learning_rate": 3.274378585086042e-06,
+      "loss": 1.8477,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9564635766203535,
+      "learning_rate": 3.2982791586998087e-06,
+      "loss": 1.9551,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.690242117872763,
+      "learning_rate": 3.322179732313576e-06,
+      "loss": 1.8477,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1414707033145675,
+      "learning_rate": 3.3460803059273427e-06,
+      "loss": 2.2051,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.568053925287825,
+      "learning_rate": 3.369980879541109e-06,
+      "loss": 1.8125,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.598501061789305,
+      "learning_rate": 3.3938814531548755e-06,
+      "loss": 2.4805,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3142814884017913,
+      "learning_rate": 3.417782026768643e-06,
+      "loss": 1.4414,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.305579896376688,
+      "learning_rate": 3.441682600382409e-06,
+      "loss": 2.0918,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.75439751499502,
+      "learning_rate": 3.4655831739961757e-06,
+      "loss": 1.168,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.850316676658855,
+      "learning_rate": 3.489483747609943e-06,
+      "loss": 1.6406,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6194294187933758,
+      "learning_rate": 3.5133843212237097e-06,
+      "loss": 1.2568,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.398771703664124,
+      "learning_rate": 3.537284894837476e-06,
+      "loss": 1.2422,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.82746065861214,
+      "learning_rate": 3.5611854684512434e-06,
+      "loss": 1.6309,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.390319596390138,
+      "learning_rate": 3.58508604206501e-06,
+      "loss": 1.5703,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2884835574107467,
+      "learning_rate": 3.608986615678776e-06,
+      "loss": 1.2715,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.287752656749102,
+      "learning_rate": 3.6328871892925436e-06,
+      "loss": 1.29,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5177653261281474,
+      "learning_rate": 3.65678776290631e-06,
+      "loss": 1.2979,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.291385985232722,
+      "learning_rate": 3.6806883365200768e-06,
+      "loss": 3.1523,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.725595848026772,
+      "learning_rate": 3.704588910133843e-06,
+      "loss": 1.502,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.498598092310823,
+      "learning_rate": 3.7284894837476104e-06,
+      "loss": 1.3555,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.368719112853617,
+      "learning_rate": 3.752390057361377e-06,
+      "loss": 1.3789,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.710464878909939,
+      "learning_rate": 3.7762906309751436e-06,
+      "loss": 1.9717,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.201079484179925,
+      "learning_rate": 3.8001912045889106e-06,
+      "loss": 1.4355,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.869099298172493,
+      "learning_rate": 3.824091778202677e-06,
+      "loss": 2.0332,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.568802567535691,
+      "learning_rate": 3.847992351816443e-06,
+      "loss": 1.2012,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.369401520196138,
+      "learning_rate": 3.871892925430211e-06,
+      "loss": 1.5723,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.454646727131855,
+      "learning_rate": 3.895793499043977e-06,
+      "loss": 1.5391,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.995957075192697,
+      "learning_rate": 3.9196940726577436e-06,
+      "loss": 1.5488,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.342878154085559,
+      "learning_rate": 3.9435946462715106e-06,
+      "loss": 1.3135,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.992826559167408,
+      "learning_rate": 3.967495219885278e-06,
+      "loss": 2.1348,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.979620756166979,
+      "learning_rate": 3.991395793499044e-06,
+      "loss": 1.3633,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.393316510301819,
+      "learning_rate": 4.015296367112811e-06,
+      "loss": 1.5098,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1721081577540424,
+      "learning_rate": 4.039196940726578e-06,
+      "loss": 1.7129,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.58204006329658,
+      "learning_rate": 4.063097514340344e-06,
+      "loss": 3.0254,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.009837392932876,
+      "learning_rate": 4.086998087954111e-06,
+      "loss": 2.1328,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.981279075432549,
+      "learning_rate": 4.110898661567878e-06,
+      "loss": 1.4648,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.205702921276771,
+      "learning_rate": 4.134799235181645e-06,
+      "loss": 2.0391,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4779320753917045,
+      "learning_rate": 4.158699808795411e-06,
+      "loss": 1.3408,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.02838228876917,
+      "learning_rate": 4.182600382409178e-06,
+      "loss": 1.5059,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.102217737498698,
+      "learning_rate": 4.206500956022945e-06,
+      "loss": 2.4375,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5854561978508395,
+      "learning_rate": 4.230401529636711e-06,
+      "loss": 1.1914,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.340014330249079,
+      "learning_rate": 4.2543021032504776e-06,
+      "loss": 1.4434,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.852777192814703,
+      "learning_rate": 4.2782026768642454e-06,
+      "loss": 1.376,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.714663528258492,
+      "learning_rate": 4.302103250478012e-06,
+      "loss": 1.584,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.679477273232807,
+      "learning_rate": 4.326003824091778e-06,
+      "loss": 1.5254,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.358184323702387,
+      "learning_rate": 4.349904397705546e-06,
+      "loss": 1.5898,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1870638492108245,
+      "learning_rate": 4.373804971319312e-06,
+      "loss": 1.1201,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.345983575922428,
+      "learning_rate": 4.397705544933078e-06,
+      "loss": 1.8047,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.947915531458824,
+      "learning_rate": 4.421606118546846e-06,
+      "loss": 2.2812,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.426318719442372,
+      "learning_rate": 4.445506692160612e-06,
+      "loss": 1.1973,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0579189369729303,
+      "learning_rate": 4.469407265774379e-06,
+      "loss": 1.75,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8817692358656117,
+      "learning_rate": 4.493307839388146e-06,
+      "loss": 1.6113,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.518283747323977,
+      "learning_rate": 4.517208413001912e-06,
+      "loss": 2.3398,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.67047270077261,
+      "learning_rate": 4.541108986615679e-06,
+      "loss": 1.6289,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.60475622805261,
+      "learning_rate": 4.565009560229445e-06,
+      "loss": 1.3975,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0265103651704655,
+      "learning_rate": 4.5889101338432124e-06,
+      "loss": 1.3594,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.179653548618808,
+      "learning_rate": 4.6128107074569795e-06,
+      "loss": 1.5977,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.050877304189026,
+      "learning_rate": 4.636711281070746e-06,
+      "loss": 1.3867,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.263117794476113,
+      "learning_rate": 4.660611854684513e-06,
+      "loss": 1.459,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.323380827592627,
+      "learning_rate": 4.68451242829828e-06,
+      "loss": 1.4004,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.348354840916063,
+      "learning_rate": 4.708413001912046e-06,
+      "loss": 1.0518,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.162057830864968,
+      "learning_rate": 4.732313575525813e-06,
+      "loss": 1.4863,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.384341049878578,
+      "learning_rate": 4.75621414913958e-06,
+      "loss": 2.0898,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4733398588513404,
+      "learning_rate": 4.780114722753346e-06,
+      "loss": 1.3008,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.857658647350561,
+      "learning_rate": 4.804015296367113e-06,
+      "loss": 1.3125,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.442991649519985,
+      "learning_rate": 4.82791586998088e-06,
+      "loss": 1.8594,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.883773838211926,
+      "learning_rate": 4.851816443594646e-06,
+      "loss": 1.4258,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.16299877214196,
+      "learning_rate": 4.875717017208413e-06,
+      "loss": 1.0996,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.013808063304936,
+      "learning_rate": 4.89961759082218e-06,
+      "loss": 2.1738,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.62061913605539,
+      "learning_rate": 4.9235181644359465e-06,
+      "loss": 1.1152,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.325127069447801,
+      "learning_rate": 4.9474187380497135e-06,
+      "loss": 2.002,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.822786058085137,
+      "learning_rate": 4.9713193116634805e-06,
+      "loss": 1.3867,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.151613814603934,
+      "learning_rate": 4.995219885277247e-06,
+      "loss": 1.6582,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3911294166395285,
+      "learning_rate": 5.019120458891014e-06,
+      "loss": 1.5859,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.379721412012378,
+      "learning_rate": 5.043021032504781e-06,
+      "loss": 1.7578,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.79954827090387,
+      "learning_rate": 5.066921606118547e-06,
+      "loss": 1.749,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.300335925554506,
+      "learning_rate": 5.090822179732314e-06,
+      "loss": 1.3438,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8144933773637817,
+      "learning_rate": 5.11472275334608e-06,
+      "loss": 1.5234,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.680664745410809,
+      "learning_rate": 5.138623326959847e-06,
+      "loss": 2.1172,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.851941016532903,
+      "learning_rate": 5.162523900573614e-06,
+      "loss": 1.6113,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.933813155267006,
+      "learning_rate": 5.18642447418738e-06,
+      "loss": 1.7324,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.937270976917201,
+      "learning_rate": 5.210325047801147e-06,
+      "loss": 1.3105,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4276886772912185,
+      "learning_rate": 5.234225621414914e-06,
+      "loss": 1.6133,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.072151339823193,
+      "learning_rate": 5.2581261950286805e-06,
+      "loss": 1.6104,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.110929071947076,
+      "learning_rate": 5.282026768642448e-06,
+      "loss": 1.5146,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.492983327943769,
+      "learning_rate": 5.3059273422562145e-06,
+      "loss": 1.5879,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.386300155824678,
+      "learning_rate": 5.329827915869981e-06,
+      "loss": 1.6797,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.528094435984237,
+      "learning_rate": 5.3537284894837486e-06,
+      "loss": 1.1572,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.367443018741811,
+      "learning_rate": 5.377629063097515e-06,
+      "loss": 1.4297,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.687337952216296,
+      "learning_rate": 5.401529636711281e-06,
+      "loss": 1.5371,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7732905325691886,
+      "learning_rate": 5.425430210325048e-06,
+      "loss": 1.9316,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.794121066656308,
+      "learning_rate": 5.449330783938815e-06,
+      "loss": 1.4023,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.017604141247396,
+      "learning_rate": 5.473231357552581e-06,
+      "loss": 1.2188,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.161904983836419,
+      "learning_rate": 5.497131931166348e-06,
+      "loss": 2.4219,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.128645495313178,
+      "learning_rate": 5.521032504780115e-06,
+      "loss": 1.5996,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.501704482162295,
+      "learning_rate": 5.544933078393882e-06,
+      "loss": 0.918,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3661726907096137,
+      "learning_rate": 5.568833652007648e-06,
+      "loss": 0.9229,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.890058913733379,
+      "learning_rate": 5.592734225621415e-06,
+      "loss": 1.2422,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.017539345468512,
+      "learning_rate": 5.616634799235182e-06,
+      "loss": 1.3965,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1001101874454955,
+      "learning_rate": 5.6405353728489485e-06,
+      "loss": 1.084,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.124539796589646,
+      "learning_rate": 5.6644359464627156e-06,
+      "loss": 0.9912,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.249840085844822,
+      "learning_rate": 5.6883365200764826e-06,
+      "loss": 1.0938,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.398998181696934,
+      "learning_rate": 5.712237093690249e-06,
+      "loss": 1.3457,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.292417343312498,
+      "learning_rate": 5.736137667304015e-06,
+      "loss": 1.123,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.007830796236203,
+      "learning_rate": 5.760038240917783e-06,
+      "loss": 1.5859,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.158594005784127,
+      "learning_rate": 5.783938814531549e-06,
+      "loss": 1.8184,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.823352813449402,
+      "learning_rate": 5.807839388145315e-06,
+      "loss": 1.3809,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6278440649061325,
+      "learning_rate": 5.831739961759083e-06,
+      "loss": 2.2871,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.542663384285694,
+      "learning_rate": 5.855640535372849e-06,
+      "loss": 1.2617,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.536273115957556,
+      "learning_rate": 5.879541108986616e-06,
+      "loss": 1.8184,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.336093487717715,
+      "learning_rate": 5.903441682600383e-06,
+      "loss": 2.4805,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.845869408712106,
+      "learning_rate": 5.927342256214149e-06,
+      "loss": 1.8262,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.672212328635715,
+      "learning_rate": 5.951242829827916e-06,
+      "loss": 2.1094,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.766881868737535,
+      "learning_rate": 5.9751434034416826e-06,
+      "loss": 1.7832,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.134232659986859,
+      "learning_rate": 5.9990439770554496e-06,
+      "loss": 1.1709,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5886855555653847,
+      "learning_rate": 6.022944550669217e-06,
+      "loss": 1.7539,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.496773322351295,
+      "learning_rate": 6.046845124282983e-06,
+      "loss": 1.3887,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.194406596718585,
+      "learning_rate": 6.07074569789675e-06,
+      "loss": 1.0859,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.4505259268843655,
+      "learning_rate": 6.094646271510517e-06,
+      "loss": 1.7148,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.014867165130096,
+      "learning_rate": 6.118546845124283e-06,
+      "loss": 1.6152,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.402083601708426,
+      "learning_rate": 6.14244741873805e-06,
+      "loss": 1.3369,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.641984425779527,
+      "learning_rate": 6.166347992351817e-06,
+      "loss": 1.4434,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.373200618738339,
+      "learning_rate": 6.190248565965583e-06,
+      "loss": 1.8164,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.608129650054078,
+      "learning_rate": 6.21414913957935e-06,
+      "loss": 1.083,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.533779694756273,
+      "learning_rate": 6.238049713193117e-06,
+      "loss": 1.0332,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.011381408117607,
+      "learning_rate": 6.261950286806883e-06,
+      "loss": 1.0,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.783615378426976,
+      "learning_rate": 6.28585086042065e-06,
+      "loss": 1.123,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.919551967604048,
+      "learning_rate": 6.3097514340344166e-06,
+      "loss": 1.4893,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.791858228752781,
+      "learning_rate": 6.3336520076481844e-06,
+      "loss": 1.5664,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.989012512428358,
+      "learning_rate": 6.357552581261951e-06,
+      "loss": 1.7734,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.883203129911792,
+      "learning_rate": 6.381453154875718e-06,
+      "loss": 1.1826,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.132959108505672,
+      "learning_rate": 6.405353728489484e-06,
+      "loss": 1.7402,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.493860656583041,
+      "learning_rate": 6.429254302103251e-06,
+      "loss": 1.0117,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.293387826467704,
+      "learning_rate": 6.453154875717017e-06,
+      "loss": 1.3164,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.980138873026146,
+      "learning_rate": 6.477055449330784e-06,
+      "loss": 2.1465,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.815217972274239,
+      "learning_rate": 6.500956022944552e-06,
+      "loss": 1.3965,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.730328621220404,
+      "learning_rate": 6.524856596558318e-06,
+      "loss": 1.6504,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9891942218768888,
+      "learning_rate": 6.548757170172084e-06,
+      "loss": 0.9746,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7787021582393643,
+      "learning_rate": 6.572657743785851e-06,
+      "loss": 1.0859,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.331809691541112,
+      "learning_rate": 6.596558317399617e-06,
+      "loss": 1.1855,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.47438256252867,
+      "learning_rate": 6.620458891013384e-06,
+      "loss": 1.709,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.609908768088329,
+      "learning_rate": 6.644359464627152e-06,
+      "loss": 1.416,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 33.249744844533204,
+      "learning_rate": 6.6682600382409185e-06,
+      "loss": 2.2266,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.996059671102788,
+      "learning_rate": 6.6921606118546855e-06,
+      "loss": 1.7207,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.243802657102665,
+      "learning_rate": 6.716061185468452e-06,
+      "loss": 1.1484,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.021015986640607,
+      "learning_rate": 6.739961759082218e-06,
+      "loss": 1.4609,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.070928420119174,
+      "learning_rate": 6.763862332695985e-06,
+      "loss": 1.9707,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.062662081419264,
+      "learning_rate": 6.787762906309751e-06,
+      "loss": 1.3633,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.598260235391513,
+      "learning_rate": 6.811663479923519e-06,
+      "loss": 1.4531,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.132913769853426,
+      "learning_rate": 6.835564053537286e-06,
+      "loss": 1.4854,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.668705240754797,
+      "learning_rate": 6.859464627151052e-06,
+      "loss": 1.3682,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1958664398756085,
+      "learning_rate": 6.883365200764818e-06,
+      "loss": 0.8428,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.163696960093071,
+      "learning_rate": 6.907265774378585e-06,
+      "loss": 1.3652,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1550043624430275,
+      "learning_rate": 6.9311663479923514e-06,
+      "loss": 1.3643,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.019866575922459,
+      "learning_rate": 6.955066921606119e-06,
+      "loss": 1.1836,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.886542810596584,
+      "learning_rate": 6.978967495219886e-06,
+      "loss": 1.5371,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.198778146601374,
+      "learning_rate": 7.0028680688336525e-06,
+      "loss": 1.752,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.625037427112735,
+      "learning_rate": 7.0267686424474195e-06,
+      "loss": 1.8477,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.623589584154307,
+      "learning_rate": 7.050669216061186e-06,
+      "loss": 1.0732,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.31814980135472,
+      "learning_rate": 7.074569789674952e-06,
+      "loss": 1.4961,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.088046334129193,
+      "learning_rate": 7.098470363288719e-06,
+      "loss": 0.9355,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.918321011071289,
+      "learning_rate": 7.122370936902487e-06,
+      "loss": 1.6992,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.811540326514273,
+      "learning_rate": 7.146271510516253e-06,
+      "loss": 1.1309,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.849371669314683,
+      "learning_rate": 7.17017208413002e-06,
+      "loss": 1.3887,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.883334807761225,
+      "learning_rate": 7.194072657743786e-06,
+      "loss": 0.9043,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.174301822760955,
+      "learning_rate": 7.217973231357552e-06,
+      "loss": 1.752,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.927470970900527,
+      "learning_rate": 7.241873804971319e-06,
+      "loss": 0.9258,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.624881580884093,
+      "learning_rate": 7.265774378585087e-06,
+      "loss": 2.1172,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.453667644432008,
+      "learning_rate": 7.289674952198853e-06,
+      "loss": 1.709,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.9688619425040725,
+      "learning_rate": 7.31357552581262e-06,
+      "loss": 1.2773,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5725700864838315,
+      "learning_rate": 7.3374760994263865e-06,
+      "loss": 1.4238,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.620449341332836,
+      "learning_rate": 7.3613766730401535e-06,
+      "loss": 1.2168,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.960022191973861,
+      "learning_rate": 7.38527724665392e-06,
+      "loss": 1.5815,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.73661793442189,
+      "learning_rate": 7.409177820267686e-06,
+      "loss": 2.1055,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3455107144212852,
+      "learning_rate": 7.433078393881454e-06,
+      "loss": 1.5645,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.370986247199563,
+      "learning_rate": 7.456978967495221e-06,
+      "loss": 1.3564,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.7937411506315035,
+      "learning_rate": 7.480879541108987e-06,
+      "loss": 1.3887,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.378236271206255,
+      "learning_rate": 7.504780114722754e-06,
+      "loss": 1.9883,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.401285722925863,
+      "learning_rate": 7.52868068833652e-06,
+      "loss": 1.6484,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.059215369963908,
+      "learning_rate": 7.552581261950287e-06,
+      "loss": 1.207,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.315651104285314,
+      "learning_rate": 7.576481835564053e-06,
+      "loss": 1.2891,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.274550270383784,
+      "learning_rate": 7.600382409177821e-06,
+      "loss": 1.3418,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.321656506721157,
+      "learning_rate": 7.624282982791587e-06,
+      "loss": 1.5859,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.017365668376748,
+      "learning_rate": 7.648183556405354e-06,
+      "loss": 1.4375,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9062421874921873,
+      "learning_rate": 7.67208413001912e-06,
+      "loss": 0.5894,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.528722670589124,
+      "learning_rate": 7.695984703632887e-06,
+      "loss": 1.3008,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4229930984467614,
+      "learning_rate": 7.719885277246655e-06,
+      "loss": 0.8857,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.034363553529437,
+      "learning_rate": 7.743785850860422e-06,
+      "loss": 1.1387,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.04558175822478,
+      "learning_rate": 7.767686424474189e-06,
+      "loss": 1.1729,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6925437084960877,
+      "learning_rate": 7.791586998087955e-06,
+      "loss": 0.9717,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.133294347258419,
+      "learning_rate": 7.815487571701721e-06,
+      "loss": 0.957,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.59853054873554,
+      "learning_rate": 7.839388145315487e-06,
+      "loss": 1.7285,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.014265585543032,
+      "learning_rate": 7.863288718929255e-06,
+      "loss": 1.877,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.875552580430806,
+      "learning_rate": 7.887189292543021e-06,
+      "loss": 1.7051,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.782446059934465,
+      "learning_rate": 7.911089866156789e-06,
+      "loss": 0.7451,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.381723242502936,
+      "learning_rate": 7.934990439770555e-06,
+      "loss": 1.9688,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.261459608442249,
+      "learning_rate": 7.958891013384321e-06,
+      "loss": 1.8086,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.079445577612615,
+      "learning_rate": 7.982791586998088e-06,
+      "loss": 1.2842,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.116614169713994,
+      "learning_rate": 8.006692160611855e-06,
+      "loss": 1.3203,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.075826568257921,
+      "learning_rate": 8.030592734225622e-06,
+      "loss": 1.0566,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7453725874452894,
+      "learning_rate": 8.05449330783939e-06,
+      "loss": 1.1777,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3301965418370103,
+      "learning_rate": 8.078393881453156e-06,
+      "loss": 1.6406,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.095352014542235,
+      "learning_rate": 8.102294455066922e-06,
+      "loss": 1.7324,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.62738058322543,
+      "learning_rate": 8.126195028680688e-06,
+      "loss": 1.0957,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9954136542866077,
+      "learning_rate": 8.150095602294456e-06,
+      "loss": 0.916,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.315986074402026,
+      "learning_rate": 8.173996175908222e-06,
+      "loss": 1.3867,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.514668399635458,
+      "learning_rate": 8.197896749521988e-06,
+      "loss": 1.3438,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.127753190964,
+      "learning_rate": 8.221797323135756e-06,
+      "loss": 1.3086,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.700824770695061,
+      "learning_rate": 8.245697896749522e-06,
+      "loss": 1.0508,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.730726390976263,
+      "learning_rate": 8.26959847036329e-06,
+      "loss": 0.9561,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.772870065760353,
+      "learning_rate": 8.293499043977056e-06,
+      "loss": 1.0156,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.239512520496379,
+      "learning_rate": 8.317399617590822e-06,
+      "loss": 1.4805,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.568899997959395,
+      "learning_rate": 8.341300191204589e-06,
+      "loss": 1.4199,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.7076282929833635,
+      "learning_rate": 8.365200764818356e-06,
+      "loss": 1.0928,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4753177085928675,
+      "learning_rate": 8.389101338432123e-06,
+      "loss": 2.041,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.229120689605698,
+      "learning_rate": 8.41300191204589e-06,
+      "loss": 1.0186,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9768620286412544,
+      "learning_rate": 8.436902485659657e-06,
+      "loss": 1.1973,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.000867272770514,
+      "learning_rate": 8.460803059273423e-06,
+      "loss": 1.1621,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.934211234740548,
+      "learning_rate": 8.484703632887189e-06,
+      "loss": 2.0508,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.487019360658022,
+      "learning_rate": 8.508604206500955e-06,
+      "loss": 1.7148,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8557961121073414,
+      "learning_rate": 8.532504780114723e-06,
+      "loss": 1.2168,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.896207408276664,
+      "learning_rate": 8.556405353728491e-06,
+      "loss": 1.5957,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.618897398481657,
+      "learning_rate": 8.580305927342257e-06,
+      "loss": 0.9043,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7546999406866437,
+      "learning_rate": 8.604206500956023e-06,
+      "loss": 1.1992,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0803092865494115,
+      "learning_rate": 8.62810707456979e-06,
+      "loss": 1.1133,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.745963960656054,
+      "learning_rate": 8.652007648183556e-06,
+      "loss": 1.6484,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.401317949491223,
+      "learning_rate": 8.675908221797323e-06,
+      "loss": 1.5215,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.441659323126067,
+      "learning_rate": 8.699808795411091e-06,
+      "loss": 1.1113,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.265480945383207,
+      "learning_rate": 8.723709369024857e-06,
+      "loss": 1.6187,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.506201109306482,
+      "learning_rate": 8.747609942638624e-06,
+      "loss": 1.1875,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.534586552759554,
+      "learning_rate": 8.77151051625239e-06,
+      "loss": 1.1455,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.765749474205371,
+      "learning_rate": 8.795411089866156e-06,
+      "loss": 1.4619,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.69676148054714,
+      "learning_rate": 8.819311663479924e-06,
+      "loss": 1.0967,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.228652833235752,
+      "learning_rate": 8.843212237093692e-06,
+      "loss": 1.6934,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.798619257247999,
+      "learning_rate": 8.867112810707458e-06,
+      "loss": 1.5312,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.662352384219652,
+      "learning_rate": 8.891013384321224e-06,
+      "loss": 1.3711,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.950236286060623,
+      "learning_rate": 8.91491395793499e-06,
+      "loss": 1.4785,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.257343605751903,
+      "learning_rate": 8.938814531548758e-06,
+      "loss": 1.3359,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.9620144372437025,
+      "learning_rate": 8.962715105162524e-06,
+      "loss": 1.1836,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.44449783663998,
+      "learning_rate": 8.986615678776292e-06,
+      "loss": 1.2441,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.02277945816629,
+      "learning_rate": 9.010516252390058e-06,
+      "loss": 1.0508,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.168591913477337,
+      "learning_rate": 9.034416826003824e-06,
+      "loss": 1.1016,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.230864605526385,
+      "learning_rate": 9.05831739961759e-06,
+      "loss": 1.0508,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9027091164709824,
+      "learning_rate": 9.082217973231358e-06,
+      "loss": 1.2129,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.922613888202125,
+      "learning_rate": 9.106118546845125e-06,
+      "loss": 1.4922,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.43158184732851,
+      "learning_rate": 9.13001912045889e-06,
+      "loss": 1.8965,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.32706024102723,
+      "learning_rate": 9.153919694072659e-06,
+      "loss": 1.6816,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.76550877567887,
+      "learning_rate": 9.177820267686425e-06,
+      "loss": 1.8789,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.560995353037955,
+      "learning_rate": 9.201720841300191e-06,
+      "loss": 1.5332,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9295040262213554,
+      "learning_rate": 9.225621414913959e-06,
+      "loss": 0.9971,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.76201144081513,
+      "learning_rate": 9.249521988527725e-06,
+      "loss": 1.3203,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1083243430436305,
+      "learning_rate": 9.273422562141491e-06,
+      "loss": 1.375,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.752706058054462,
+      "learning_rate": 9.297323135755259e-06,
+      "loss": 1.0039,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.60479538272032,
+      "learning_rate": 9.321223709369025e-06,
+      "loss": 1.0879,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.656454238460029,
+      "learning_rate": 9.345124282982791e-06,
+      "loss": 1.1758,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.51222875221749,
+      "learning_rate": 9.36902485659656e-06,
+      "loss": 0.7773,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.72828503785743,
+      "learning_rate": 9.392925430210326e-06,
+      "loss": 1.4395,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.67530424223932,
+      "learning_rate": 9.416826003824092e-06,
+      "loss": 0.814,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.340855342886432,
+      "learning_rate": 9.440726577437858e-06,
+      "loss": 1.7197,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.69033158291245,
+      "learning_rate": 9.464627151051626e-06,
+      "loss": 1.3633,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.84720041121181,
+      "learning_rate": 9.488527724665394e-06,
+      "loss": 1.2793,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.8224230431981345,
+      "learning_rate": 9.51242829827916e-06,
+      "loss": 1.2559,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.373789374563205,
+      "learning_rate": 9.536328871892926e-06,
+      "loss": 1.2129,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.117685634979921,
+      "learning_rate": 9.560229445506692e-06,
+      "loss": 1.0645,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.111271061807083,
+      "learning_rate": 9.584130019120458e-06,
+      "loss": 1.3711,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.878893757682663,
+      "learning_rate": 9.608030592734226e-06,
+      "loss": 2.5117,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.127546175564769,
+      "learning_rate": 9.631931166347994e-06,
+      "loss": 1.2524,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.22007832278206,
+      "learning_rate": 9.65583173996176e-06,
+      "loss": 0.6768,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.7132281734031745,
+      "learning_rate": 9.679732313575526e-06,
+      "loss": 1.0312,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.581416042469058,
+      "learning_rate": 9.703632887189293e-06,
+      "loss": 1.5596,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7973156583040697,
+      "learning_rate": 9.727533460803059e-06,
+      "loss": 1.2217,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 31.55219993599149,
+      "learning_rate": 9.751434034416827e-06,
+      "loss": 1.6191,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.570309931900267,
+      "learning_rate": 9.775334608030594e-06,
+      "loss": 1.9551,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.434676741854307,
+      "learning_rate": 9.79923518164436e-06,
+      "loss": 1.5117,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.506824594192742,
+      "learning_rate": 9.823135755258127e-06,
+      "loss": 1.0684,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.22334247016146,
+      "learning_rate": 9.847036328871893e-06,
+      "loss": 1.1152,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.189833986860409,
+      "learning_rate": 9.870936902485659e-06,
+      "loss": 1.4414,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.088603034636933,
+      "learning_rate": 9.894837476099427e-06,
+      "loss": 1.7754,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.250083491224274,
+      "learning_rate": 9.918738049713193e-06,
+      "loss": 1.4316,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6471088979072555,
+      "learning_rate": 9.942638623326961e-06,
+      "loss": 0.9824,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.09304657487768,
+      "learning_rate": 9.966539196940727e-06,
+      "loss": 1.1895,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 24.529856648327563,
+      "learning_rate": 9.990439770554493e-06,
+      "loss": 1.8271,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.92663499057138,
+      "learning_rate": 1.0014340344168261e-05,
+      "loss": 1.9023,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.251657859982913,
+      "learning_rate": 1.0038240917782027e-05,
+      "loss": 0.9375,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.583478363769167,
+      "learning_rate": 1.0062141491395794e-05,
+      "loss": 1.0693,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.387458733845366,
+      "learning_rate": 1.0086042065009561e-05,
+      "loss": 1.0586,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.836094214873508,
+      "learning_rate": 1.0109942638623328e-05,
+      "loss": 1.5859,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.096553361188105,
+      "learning_rate": 1.0133843212237094e-05,
+      "loss": 0.7793,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3492815761780514,
+      "learning_rate": 1.0157743785850862e-05,
+      "loss": 1.5625,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.443291980944949,
+      "learning_rate": 1.0181644359464628e-05,
+      "loss": 1.8457,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.4015462239404375,
+      "learning_rate": 1.0205544933078394e-05,
+      "loss": 1.0723,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.552129237057759,
+      "learning_rate": 1.022944550669216e-05,
+      "loss": 1.6777,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.92814542268893,
+      "learning_rate": 1.0253346080305928e-05,
+      "loss": 0.8359,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.332454151901733,
+      "learning_rate": 1.0277246653919694e-05,
+      "loss": 0.9688,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.941344933629197,
+      "learning_rate": 1.0301147227533462e-05,
+      "loss": 1.0264,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.747800680657284,
+      "learning_rate": 1.0325047801147228e-05,
+      "loss": 1.0117,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.880032595589527,
+      "learning_rate": 1.0348948374760994e-05,
+      "loss": 1.3877,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.315843634741467,
+      "learning_rate": 1.037284894837476e-05,
+      "loss": 1.4473,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.722057615873419,
+      "learning_rate": 1.0396749521988528e-05,
+      "loss": 1.4492,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0415729272288035,
+      "learning_rate": 1.0420650095602295e-05,
+      "loss": 1.3125,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.446429740950581,
+      "learning_rate": 1.0444550669216062e-05,
+      "loss": 1.0645,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.274013222473569,
+      "learning_rate": 1.0468451242829829e-05,
+      "loss": 1.334,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.070579392472443,
+      "learning_rate": 1.0492351816443595e-05,
+      "loss": 1.2402,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.167350232017576,
+      "learning_rate": 1.0516252390057361e-05,
+      "loss": 1.7246,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.061130668154767,
+      "learning_rate": 1.0540152963671129e-05,
+      "loss": 1.1592,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.430322931006016,
+      "learning_rate": 1.0564053537284897e-05,
+      "loss": 1.1538,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0815251816983675,
+      "learning_rate": 1.0587954110898663e-05,
+      "loss": 1.1348,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.993086807128953,
+      "learning_rate": 1.0611854684512429e-05,
+      "loss": 1.1514,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7662402398318164,
+      "learning_rate": 1.0635755258126195e-05,
+      "loss": 0.7744,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.288647796649435,
+      "learning_rate": 1.0659655831739961e-05,
+      "loss": 1.8496,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.944050129267152,
+      "learning_rate": 1.068355640535373e-05,
+      "loss": 0.9971,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.5333365433873185,
+      "learning_rate": 1.0707456978967497e-05,
+      "loss": 0.9121,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.404537225271846,
+      "learning_rate": 1.0731357552581263e-05,
+      "loss": 0.9092,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.198762693258202,
+      "learning_rate": 1.075525812619503e-05,
+      "loss": 1.3662,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.87513408454078,
+      "learning_rate": 1.0779158699808796e-05,
+      "loss": 1.5586,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.776590979259382,
+      "learning_rate": 1.0803059273422562e-05,
+      "loss": 1.7363,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.574838477746015,
+      "learning_rate": 1.082695984703633e-05,
+      "loss": 1.001,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.05314985498519,
+      "learning_rate": 1.0850860420650096e-05,
+      "loss": 1.9053,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.1793080186569345,
+      "learning_rate": 1.0874760994263864e-05,
+      "loss": 1.2891,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7085919807900773,
+      "learning_rate": 1.089866156787763e-05,
+      "loss": 1.4395,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.114881736136502,
+      "learning_rate": 1.0922562141491396e-05,
+      "loss": 1.3008,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.770777412557744,
+      "learning_rate": 1.0946462715105162e-05,
+      "loss": 1.0781,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.6089839533350005,
+      "learning_rate": 1.097036328871893e-05,
+      "loss": 2.3613,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.01124169756921,
+      "learning_rate": 1.0994263862332696e-05,
+      "loss": 1.3672,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.174306038228988,
+      "learning_rate": 1.1018164435946464e-05,
+      "loss": 1.459,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.448593886172002,
+      "learning_rate": 1.104206500956023e-05,
+      "loss": 1.3164,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.660623787216551,
+      "learning_rate": 1.1065965583173996e-05,
+      "loss": 1.3154,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.872817009103133,
+      "learning_rate": 1.1089866156787764e-05,
+      "loss": 1.6934,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.865464175278362,
+      "learning_rate": 1.111376673040153e-05,
+      "loss": 1.7344,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.390249753413547,
+      "learning_rate": 1.1137667304015297e-05,
+      "loss": 1.1797,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.157183590423085,
+      "learning_rate": 1.1161567877629063e-05,
+      "loss": 1.2031,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.766010820159787,
+      "learning_rate": 1.118546845124283e-05,
+      "loss": 1.5527,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.540222544992331,
+      "learning_rate": 1.1209369024856597e-05,
+      "loss": 1.4219,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.338283344439134,
+      "learning_rate": 1.1233269598470365e-05,
+      "loss": 1.0957,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.728811133049836,
+      "learning_rate": 1.1257170172084131e-05,
+      "loss": 1.8438,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.341128414528012,
+      "learning_rate": 1.1281070745697897e-05,
+      "loss": 1.5723,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0998978074980394,
+      "learning_rate": 1.1304971319311663e-05,
+      "loss": 1.4414,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.666969811720528,
+      "learning_rate": 1.1328871892925431e-05,
+      "loss": 1.2559,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.088207335398286,
+      "learning_rate": 1.1352772466539197e-05,
+      "loss": 1.1064,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.087744771508419,
+      "learning_rate": 1.1376673040152965e-05,
+      "loss": 1.252,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.768569586904586,
+      "learning_rate": 1.1400573613766731e-05,
+      "loss": 0.915,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.899938070700793,
+      "learning_rate": 1.1424474187380497e-05,
+      "loss": 0.9795,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.541800717995464,
+      "learning_rate": 1.1448374760994264e-05,
+      "loss": 1.3135,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.973704753675,
+      "learning_rate": 1.147227533460803e-05,
+      "loss": 0.9941,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.756030834016329,
+      "learning_rate": 1.1496175908221798e-05,
+      "loss": 1.6875,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.432955732684855,
+      "learning_rate": 1.1520076481835566e-05,
+      "loss": 2.377,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.4747543141487105,
+      "learning_rate": 1.1543977055449332e-05,
+      "loss": 1.1992,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.055393563390309,
+      "learning_rate": 1.1567877629063098e-05,
+      "loss": 1.4912,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.83307953415409,
+      "learning_rate": 1.1591778202676864e-05,
+      "loss": 1.2422,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.642938113192135,
+      "learning_rate": 1.161567877629063e-05,
+      "loss": 0.7637,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.632577414785553,
+      "learning_rate": 1.1639579349904398e-05,
+      "loss": 0.748,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.387252726654971,
+      "learning_rate": 1.1663479923518166e-05,
+      "loss": 1.5684,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.502892706307783,
+      "learning_rate": 1.1687380497131932e-05,
+      "loss": 1.2949,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.476449516563786,
+      "learning_rate": 1.1711281070745698e-05,
+      "loss": 1.5215,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.472293599411419,
+      "learning_rate": 1.1735181644359464e-05,
+      "loss": 0.729,
+      "step": 491
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.331173310940403,
+      "learning_rate": 1.1759082217973232e-05,
+      "loss": 1.3896,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.127005394078798,
+      "learning_rate": 1.1782982791586999e-05,
+      "loss": 1.2246,
+      "step": 493
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.498241124197879,
+      "learning_rate": 1.1806883365200766e-05,
+      "loss": 1.5527,
+      "step": 494
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.482739145071026,
+      "learning_rate": 1.1830783938814533e-05,
+      "loss": 0.9775,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.087462277486246,
+      "learning_rate": 1.1854684512428299e-05,
+      "loss": 1.1421,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4388246238207607,
+      "learning_rate": 1.1878585086042065e-05,
+      "loss": 1.4531,
+      "step": 497
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.095839944576186,
+      "learning_rate": 1.1902485659655833e-05,
+      "loss": 1.4785,
+      "step": 498
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6525688224433974,
+      "learning_rate": 1.1926386233269599e-05,
+      "loss": 1.3125,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.230025760603499,
+      "learning_rate": 1.1950286806883365e-05,
+      "loss": 1.1719,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.405249780759233,
+      "learning_rate": 1.1974187380497133e-05,
+      "loss": 1.3398,
+      "step": 501
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.527536347298206,
+      "learning_rate": 1.1998087954110899e-05,
+      "loss": 0.9082,
+      "step": 502
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.78044529472477,
+      "learning_rate": 1.2021988527724665e-05,
+      "loss": 1.5312,
+      "step": 503
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.937155023426123,
+      "learning_rate": 1.2045889101338433e-05,
+      "loss": 1.417,
+      "step": 504
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.260251708009438,
+      "learning_rate": 1.20697896749522e-05,
+      "loss": 1.2021,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.839725914752988,
+      "learning_rate": 1.2093690248565966e-05,
+      "loss": 0.9697,
+      "step": 506
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.168099398652328,
+      "learning_rate": 1.2117590822179733e-05,
+      "loss": 1.3262,
+      "step": 507
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.759771635844584,
+      "learning_rate": 1.21414913957935e-05,
+      "loss": 1.2832,
+      "step": 508
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.332020261947747,
+      "learning_rate": 1.2165391969407266e-05,
+      "loss": 1.293,
+      "step": 509
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.157173996621846,
+      "learning_rate": 1.2189292543021034e-05,
+      "loss": 1.5137,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.875204424362253,
+      "learning_rate": 1.22131931166348e-05,
+      "loss": 1.2256,
+      "step": 511
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.172060158724653,
+      "learning_rate": 1.2237093690248566e-05,
+      "loss": 0.9736,
+      "step": 512
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.710436839481291,
+      "learning_rate": 1.2260994263862332e-05,
+      "loss": 0.9736,
+      "step": 513
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.407809170487996,
+      "learning_rate": 1.22848948374761e-05,
+      "loss": 1.2793,
+      "step": 514
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.766244066847275,
+      "learning_rate": 1.2308795411089868e-05,
+      "loss": 1.4414,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.753744141738633,
+      "learning_rate": 1.2332695984703634e-05,
+      "loss": 1.3477,
+      "step": 516
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.316333338609315,
+      "learning_rate": 1.23565965583174e-05,
+      "loss": 1.4082,
+      "step": 517
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.121745235262178,
+      "learning_rate": 1.2380497131931166e-05,
+      "loss": 1.6719,
+      "step": 518
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.307789321072141,
+      "learning_rate": 1.2404397705544933e-05,
+      "loss": 1.7246,
+      "step": 519
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.68321025934211,
+      "learning_rate": 1.24282982791587e-05,
+      "loss": 1.8184,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2701738298040235,
+      "learning_rate": 1.2452198852772468e-05,
+      "loss": 1.2734,
+      "step": 521
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.824960086945186,
+      "learning_rate": 1.2476099426386234e-05,
+      "loss": 2.0605,
+      "step": 522
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.548651731634303,
+      "learning_rate": 1.25e-05,
+      "loss": 1.2881,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.784631667674378,
+      "learning_rate": 1.2523900573613767e-05,
+      "loss": 1.9941,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.382372443744345,
+      "learning_rate": 1.2547801147227533e-05,
+      "loss": 1.0859,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.85481686880857,
+      "learning_rate": 1.25717017208413e-05,
+      "loss": 1.168,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.213935321449134,
+      "learning_rate": 1.2595602294455067e-05,
+      "loss": 0.856,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2500249076562175,
+      "learning_rate": 1.2619502868068833e-05,
+      "loss": 1.6914,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.709378519271928,
+      "learning_rate": 1.26434034416826e-05,
+      "loss": 1.0684,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.598122555931406,
+      "learning_rate": 1.2667304015296369e-05,
+      "loss": 1.2139,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.385272352516207,
+      "learning_rate": 1.2691204588910135e-05,
+      "loss": 1.4473,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.707745044099267,
+      "learning_rate": 1.2715105162523901e-05,
+      "loss": 1.4219,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.19959559490112,
+      "learning_rate": 1.2739005736137669e-05,
+      "loss": 1.4102,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.274595684448185,
+      "learning_rate": 1.2762906309751435e-05,
+      "loss": 1.499,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.350695106769806,
+      "learning_rate": 1.2786806883365201e-05,
+      "loss": 0.9521,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.263179315484696,
+      "learning_rate": 1.2810707456978968e-05,
+      "loss": 1.2725,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.425024592133596,
+      "learning_rate": 1.2834608030592735e-05,
+      "loss": 1.6582,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.990912571637525,
+      "learning_rate": 1.2858508604206502e-05,
+      "loss": 1.5508,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.881196670209358,
+      "learning_rate": 1.2882409177820268e-05,
+      "loss": 1.2285,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.183763146283049,
+      "learning_rate": 1.2906309751434034e-05,
+      "loss": 0.543,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.453584215344017,
+      "learning_rate": 1.29302103250478e-05,
+      "loss": 1.1855,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.001698790426621,
+      "learning_rate": 1.2954110898661568e-05,
+      "loss": 1.7773,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.881899988429737,
+      "learning_rate": 1.2978011472275336e-05,
+      "loss": 1.3643,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.610360329261301,
+      "learning_rate": 1.3001912045889104e-05,
+      "loss": 1.4297,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.515350940152584,
+      "learning_rate": 1.302581261950287e-05,
+      "loss": 1.7754,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.909014808655479,
+      "learning_rate": 1.3049713193116636e-05,
+      "loss": 0.8818,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6421809330197394,
+      "learning_rate": 1.3073613766730402e-05,
+      "loss": 1.377,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.088726659910502,
+      "learning_rate": 1.3097514340344168e-05,
+      "loss": 1.1445,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.267738348334943,
+      "learning_rate": 1.3121414913957936e-05,
+      "loss": 1.2266,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.73284243027159,
+      "learning_rate": 1.3145315487571702e-05,
+      "loss": 0.709,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.78191038511124,
+      "learning_rate": 1.3169216061185469e-05,
+      "loss": 1.7051,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.646192526345727,
+      "learning_rate": 1.3193116634799235e-05,
+      "loss": 1.7344,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.822288884241532,
+      "learning_rate": 1.3217017208413001e-05,
+      "loss": 1.1367,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.72651886760748,
+      "learning_rate": 1.3240917782026769e-05,
+      "loss": 1.2715,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.004227171055042,
+      "learning_rate": 1.3264818355640535e-05,
+      "loss": 1.127,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.074658146511455,
+      "learning_rate": 1.3288718929254305e-05,
+      "loss": 1.4492,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.82908970107375,
+      "learning_rate": 1.331261950286807e-05,
+      "loss": 1.1592,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5912978057388103,
+      "learning_rate": 1.3336520076481837e-05,
+      "loss": 1.1958,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.828413641687872,
+      "learning_rate": 1.3360420650095603e-05,
+      "loss": 0.8779,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9697497452813235,
+      "learning_rate": 1.3384321223709371e-05,
+      "loss": 1.2393,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.024959180162892,
+      "learning_rate": 1.3408221797323137e-05,
+      "loss": 0.8799,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.894116405401018,
+      "learning_rate": 1.3432122370936903e-05,
+      "loss": 1.2637,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.895057411757047,
+      "learning_rate": 1.345602294455067e-05,
+      "loss": 1.4473,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.82548918408373,
+      "learning_rate": 1.3479923518164436e-05,
+      "loss": 1.4746,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.865573474608764,
+      "learning_rate": 1.3503824091778204e-05,
+      "loss": 1.9053,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.192260760235627,
+      "learning_rate": 1.352772466539197e-05,
+      "loss": 0.6841,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.792242041448269,
+      "learning_rate": 1.3551625239005736e-05,
+      "loss": 1.2031,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.914881943002599,
+      "learning_rate": 1.3575525812619502e-05,
+      "loss": 1.8496,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9004642210288725,
+      "learning_rate": 1.3599426386233272e-05,
+      "loss": 1.1123,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.559774969805917,
+      "learning_rate": 1.3623326959847038e-05,
+      "loss": 1.1719,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.06090150342357,
+      "learning_rate": 1.3647227533460804e-05,
+      "loss": 0.9922,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.31697053197031,
+      "learning_rate": 1.3671128107074572e-05,
+      "loss": 1.5293,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9406962667338865,
+      "learning_rate": 1.3695028680688338e-05,
+      "loss": 0.9766,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.963060185961364,
+      "learning_rate": 1.3718929254302104e-05,
+      "loss": 1.2822,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.983198762692828,
+      "learning_rate": 1.374282982791587e-05,
+      "loss": 1.6934,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.525706743494888,
+      "learning_rate": 1.3766730401529636e-05,
+      "loss": 1.6152,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.085208747600863,
+      "learning_rate": 1.3790630975143404e-05,
+      "loss": 1.3047,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3560658324449033,
+      "learning_rate": 1.381453154875717e-05,
+      "loss": 1.3906,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.360205939023261,
+      "learning_rate": 1.3838432122370937e-05,
+      "loss": 1.0762,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.246807521068428,
+      "learning_rate": 1.3862332695984703e-05,
+      "loss": 1.582,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.281826676561109,
+      "learning_rate": 1.3886233269598469e-05,
+      "loss": 1.3223,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.412412050207044,
+      "learning_rate": 1.3910133843212239e-05,
+      "loss": 1.4844,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.890613815036133,
+      "learning_rate": 1.3934034416826006e-05,
+      "loss": 1.0527,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.799225348488724,
+      "learning_rate": 1.3957934990439773e-05,
+      "loss": 0.7012,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.719168739181423,
+      "learning_rate": 1.3981835564053539e-05,
+      "loss": 0.9111,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.698094241725366,
+      "learning_rate": 1.4005736137667305e-05,
+      "loss": 1.4844,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.301498981389035,
+      "learning_rate": 1.4029636711281071e-05,
+      "loss": 1.1602,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.888198982072035,
+      "learning_rate": 1.4053537284894839e-05,
+      "loss": 1.1562,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.241867614524379,
+      "learning_rate": 1.4077437858508605e-05,
+      "loss": 1.0957,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.47115501084303,
+      "learning_rate": 1.4101338432122371e-05,
+      "loss": 1.0703,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9720652993127383,
+      "learning_rate": 1.4125239005736138e-05,
+      "loss": 1.0146,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4490334967592844,
+      "learning_rate": 1.4149139579349904e-05,
+      "loss": 1.4141,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.910268865673639,
+      "learning_rate": 1.4173040152963672e-05,
+      "loss": 1.2109,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6004952196074385,
+      "learning_rate": 1.4196940726577438e-05,
+      "loss": 0.9785,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.606406287100166,
+      "learning_rate": 1.4220841300191207e-05,
+      "loss": 1.8613,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.315231924099925,
+      "learning_rate": 1.4244741873804973e-05,
+      "loss": 1.3516,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7711213706146793,
+      "learning_rate": 1.426864244741874e-05,
+      "loss": 1.0781,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.420197041712424,
+      "learning_rate": 1.4292543021032506e-05,
+      "loss": 1.4102,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.237815159167185,
+      "learning_rate": 1.4316443594646272e-05,
+      "loss": 1.1992,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.794334158001333,
+      "learning_rate": 1.434034416826004e-05,
+      "loss": 0.7886,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.363970774025489,
+      "learning_rate": 1.4364244741873806e-05,
+      "loss": 1.3184,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.716645138576556,
+      "learning_rate": 1.4388145315487572e-05,
+      "loss": 0.9824,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.960970083961777,
+      "learning_rate": 1.4412045889101338e-05,
+      "loss": 1.1943,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.450099997016608,
+      "learning_rate": 1.4435946462715105e-05,
+      "loss": 1.1611,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4760627987525745,
+      "learning_rate": 1.4459847036328872e-05,
+      "loss": 1.5117,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.314930548788606,
+      "learning_rate": 1.4483747609942639e-05,
+      "loss": 0.4482,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.154985048930493,
+      "learning_rate": 1.4507648183556405e-05,
+      "loss": 0.9131,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.159080975326901,
+      "learning_rate": 1.4531548757170174e-05,
+      "loss": 0.7285,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.857695686411134,
+      "learning_rate": 1.455544933078394e-05,
+      "loss": 1.3516,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.799872679792776,
+      "learning_rate": 1.4579349904397707e-05,
+      "loss": 1.5977,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7447686899865933,
+      "learning_rate": 1.4603250478011474e-05,
+      "loss": 1.0029,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.431454685160159,
+      "learning_rate": 1.462715105162524e-05,
+      "loss": 0.9141,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.697479580403056,
+      "learning_rate": 1.4651051625239007e-05,
+      "loss": 0.9111,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 20.87694864828865,
+      "learning_rate": 1.4674952198852773e-05,
+      "loss": 2.4609,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.7535798203486355,
+      "learning_rate": 1.469885277246654e-05,
+      "loss": 1.6348,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.515632025625709,
+      "learning_rate": 1.4722753346080307e-05,
+      "loss": 1.4219,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5172599076430213,
+      "learning_rate": 1.4746653919694073e-05,
+      "loss": 1.3984,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2461986370871285,
+      "learning_rate": 1.477055449330784e-05,
+      "loss": 1.2812,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.310531251413667,
+      "learning_rate": 1.4794455066921606e-05,
+      "loss": 0.9429,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.103290278454406,
+      "learning_rate": 1.4818355640535372e-05,
+      "loss": 1.2285,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.660740629626396,
+      "learning_rate": 1.484225621414914e-05,
+      "loss": 1.5645,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.874211956789224,
+      "learning_rate": 1.4866156787762907e-05,
+      "loss": 1.2402,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.621489558677678,
+      "learning_rate": 1.4890057361376675e-05,
+      "loss": 1.2959,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.862840446558873,
+      "learning_rate": 1.4913957934990441e-05,
+      "loss": 1.6152,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.311136936951805,
+      "learning_rate": 1.4937858508604208e-05,
+      "loss": 1.1885,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.033648185903365,
+      "learning_rate": 1.4961759082217974e-05,
+      "loss": 1.2148,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.260502124977007,
+      "learning_rate": 1.498565965583174e-05,
+      "loss": 1.3301,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.587380474954199,
+      "learning_rate": 1.5009560229445508e-05,
+      "loss": 1.1104,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.370965915428887,
+      "learning_rate": 1.5033460803059274e-05,
+      "loss": 0.8057,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.144451863268319,
+      "learning_rate": 1.505736137667304e-05,
+      "loss": 1.3525,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8346218486690495,
+      "learning_rate": 1.5081261950286806e-05,
+      "loss": 1.2793,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.941345575688715,
+      "learning_rate": 1.5105162523900574e-05,
+      "loss": 1.3984,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.417263254623221,
+      "learning_rate": 1.512906309751434e-05,
+      "loss": 1.6201,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.984594511584685,
+      "learning_rate": 1.5152963671128107e-05,
+      "loss": 0.8481,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.244775814970722,
+      "learning_rate": 1.5176864244741876e-05,
+      "loss": 0.8379,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.40070096414538,
+      "learning_rate": 1.5200764818355642e-05,
+      "loss": 1.3447,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6497839458743995,
+      "learning_rate": 1.5224665391969408e-05,
+      "loss": 1.2461,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.240732523061708,
+      "learning_rate": 1.5248565965583175e-05,
+      "loss": 0.9082,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.270331231282695,
+      "learning_rate": 1.5272466539196943e-05,
+      "loss": 1.2773,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.168887279737335,
+      "learning_rate": 1.529636711281071e-05,
+      "loss": 1.2285,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.816302562477016,
+      "learning_rate": 1.5320267686424475e-05,
+      "loss": 1.4336,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.806141699718497,
+      "learning_rate": 1.534416826003824e-05,
+      "loss": 0.4824,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.173446759235446,
+      "learning_rate": 1.5368068833652007e-05,
+      "loss": 0.9326,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.087444778675986,
+      "learning_rate": 1.5391969407265773e-05,
+      "loss": 1.0713,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3254433874055587,
+      "learning_rate": 1.541586998087954e-05,
+      "loss": 1.5039,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8407944892502104,
+      "learning_rate": 1.543977055449331e-05,
+      "loss": 1.5859,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.054362786312354,
+      "learning_rate": 1.5463671128107075e-05,
+      "loss": 1.1875,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.72235832609373,
+      "learning_rate": 1.5487571701720845e-05,
+      "loss": 1.1729,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.031130013308683,
+      "learning_rate": 1.551147227533461e-05,
+      "loss": 1.0254,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.19369663255735,
+      "learning_rate": 1.5535372848948377e-05,
+      "loss": 1.4668,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.023351309646475,
+      "learning_rate": 1.5559273422562143e-05,
+      "loss": 0.876,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.088749697826135,
+      "learning_rate": 1.558317399617591e-05,
+      "loss": 1.9062,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.26328176897216,
+      "learning_rate": 1.5607074569789676e-05,
+      "loss": 1.3721,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.960149843659727,
+      "learning_rate": 1.5630975143403442e-05,
+      "loss": 1.4668,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.004921984347262,
+      "learning_rate": 1.5654875717017208e-05,
+      "loss": 1.3516,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.153224860039078,
+      "learning_rate": 1.5678776290630974e-05,
+      "loss": 1.0293,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.112258885512863,
+      "learning_rate": 1.570267686424474e-05,
+      "loss": 1.377,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9338320647614227,
+      "learning_rate": 1.572657743785851e-05,
+      "loss": 1.2422,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9301903865267773,
+      "learning_rate": 1.5750478011472276e-05,
+      "loss": 0.8662,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8672395336619343,
+      "learning_rate": 1.5774378585086042e-05,
+      "loss": 1.043,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.436836725231954,
+      "learning_rate": 1.5798279158699812e-05,
+      "loss": 1.1562,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.224791535293841,
+      "learning_rate": 1.5822179732313578e-05,
+      "loss": 0.7793,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.323070903525608,
+      "learning_rate": 1.5846080305927344e-05,
+      "loss": 2.0195,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.053239809285661,
+      "learning_rate": 1.586998087954111e-05,
+      "loss": 1.4473,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.806630445699927,
+      "learning_rate": 1.5893881453154877e-05,
+      "loss": 1.4004,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.917980842616342,
+      "learning_rate": 1.5917782026768643e-05,
+      "loss": 0.7729,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.735184502717954,
+      "learning_rate": 1.594168260038241e-05,
+      "loss": 0.998,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.947602211062051,
+      "learning_rate": 1.5965583173996175e-05,
+      "loss": 1.3945,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.252783593034634,
+      "learning_rate": 1.5989483747609945e-05,
+      "loss": 1.2383,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.866151710951112,
+      "learning_rate": 1.601338432122371e-05,
+      "loss": 1.1138,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.011151758613975,
+      "learning_rate": 1.6037284894837477e-05,
+      "loss": 1.6484,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0641381689646208,
+      "learning_rate": 1.6061185468451243e-05,
+      "loss": 0.8057,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.884314036608053,
+      "learning_rate": 1.608508604206501e-05,
+      "loss": 1.0303,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.573593657251417,
+      "learning_rate": 1.610898661567878e-05,
+      "loss": 1.0469,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.175529979160558,
+      "learning_rate": 1.6132887189292545e-05,
+      "loss": 1.9844,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.112814791009904,
+      "learning_rate": 1.615678776290631e-05,
+      "loss": 1.791,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.331437409696903,
+      "learning_rate": 1.6180688336520077e-05,
+      "loss": 0.8398,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.58087280955895,
+      "learning_rate": 1.6204588910133844e-05,
+      "loss": 1.2402,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.455210414930881,
+      "learning_rate": 1.622848948374761e-05,
+      "loss": 1.2832,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2625637764378963,
+      "learning_rate": 1.6252390057361376e-05,
+      "loss": 1.1875,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.302898257552046,
+      "learning_rate": 1.6276290630975145e-05,
+      "loss": 1.0449,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.775168705810145,
+      "learning_rate": 1.630019120458891e-05,
+      "loss": 1.0181,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7929081543659944,
+      "learning_rate": 1.6324091778202678e-05,
+      "loss": 0.7334,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.329540992428358,
+      "learning_rate": 1.6347992351816444e-05,
+      "loss": 0.9668,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.147061534508436,
+      "learning_rate": 1.637189292543021e-05,
+      "loss": 1.832,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0239271736304625,
+      "learning_rate": 1.6395793499043976e-05,
+      "loss": 1.1895,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.810688619194213,
+      "learning_rate": 1.6419694072657746e-05,
+      "loss": 0.6895,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.265725543239142,
+      "learning_rate": 1.6443594646271512e-05,
+      "loss": 1.3691,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0624135842100744,
+      "learning_rate": 1.6467495219885278e-05,
+      "loss": 0.731,
+      "step": 689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.461355028658788,
+      "learning_rate": 1.6491395793499044e-05,
+      "loss": 0.8501,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.530791638436236,
+      "learning_rate": 1.651529636711281e-05,
+      "loss": 2.2031,
+      "step": 691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.790665501114771,
+      "learning_rate": 1.653919694072658e-05,
+      "loss": 1.1162,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.280877981242073,
+      "learning_rate": 1.6563097514340346e-05,
+      "loss": 1.1387,
+      "step": 693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.219415456775382,
+      "learning_rate": 1.6586998087954112e-05,
+      "loss": 0.8691,
+      "step": 694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.293968976709312,
+      "learning_rate": 1.661089866156788e-05,
+      "loss": 1.7598,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.8341435914204665,
+      "learning_rate": 1.6634799235181645e-05,
+      "loss": 0.8057,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.3685726364434885,
+      "learning_rate": 1.665869980879541e-05,
+      "loss": 1.4629,
+      "step": 697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.79909931022353,
+      "learning_rate": 1.6682600382409177e-05,
+      "loss": 3.0742,
+      "step": 698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.603133394916076,
+      "learning_rate": 1.6706500956022943e-05,
+      "loss": 1.291,
+      "step": 699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.013969378959008,
+      "learning_rate": 1.6730401529636713e-05,
+      "loss": 0.7998,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.75861467129092,
+      "learning_rate": 1.675430210325048e-05,
+      "loss": 1.1626,
+      "step": 701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.136117694134747,
+      "learning_rate": 1.6778202676864245e-05,
+      "loss": 1.6523,
+      "step": 702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.643367364866809,
+      "learning_rate": 1.680210325047801e-05,
+      "loss": 1.4824,
+      "step": 703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8998913089571725,
+      "learning_rate": 1.682600382409178e-05,
+      "loss": 1.5488,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.347474239504654,
+      "learning_rate": 1.6849904397705547e-05,
+      "loss": 1.2012,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.667965317441999,
+      "learning_rate": 1.6873804971319313e-05,
+      "loss": 1.4941,
+      "step": 706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.0005427558835125,
+      "learning_rate": 1.689770554493308e-05,
+      "loss": 1.043,
+      "step": 707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.034663747273157,
+      "learning_rate": 1.6921606118546846e-05,
+      "loss": 2.2227,
+      "step": 708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4057786157787726,
+      "learning_rate": 1.6945506692160612e-05,
+      "loss": 1.3047,
+      "step": 709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.117753303788671,
+      "learning_rate": 1.6969407265774378e-05,
+      "loss": 1.0527,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.741528599303322,
+      "learning_rate": 1.6993307839388144e-05,
+      "loss": 1.0752,
+      "step": 711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.702834415013373,
+      "learning_rate": 1.701720841300191e-05,
+      "loss": 1.8242,
+      "step": 712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.37587194838328,
+      "learning_rate": 1.704110898661568e-05,
+      "loss": 1.1709,
+      "step": 713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.935638135815365,
+      "learning_rate": 1.7065009560229446e-05,
+      "loss": 1.4082,
+      "step": 714
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.008063741920802,
+      "learning_rate": 1.7088910133843216e-05,
+      "loss": 1.2207,
+      "step": 715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.249491349057366,
+      "learning_rate": 1.7112810707456982e-05,
+      "loss": 1.7812,
+      "step": 716
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.724109561718806,
+      "learning_rate": 1.7136711281070748e-05,
+      "loss": 0.7939,
+      "step": 717
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.133455586388848,
+      "learning_rate": 1.7160611854684514e-05,
+      "loss": 2.2207,
+      "step": 718
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.837035416563926,
+      "learning_rate": 1.718451242829828e-05,
+      "loss": 1.4238,
+      "step": 719
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2970714177408125,
+      "learning_rate": 1.7208413001912046e-05,
+      "loss": 1.2617,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.182276333398534,
+      "learning_rate": 1.7232313575525813e-05,
+      "loss": 1.377,
+      "step": 721
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9102865100356707,
+      "learning_rate": 1.725621414913958e-05,
+      "loss": 1.1494,
+      "step": 722
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.121055691167921,
+      "learning_rate": 1.7280114722753345e-05,
+      "loss": 1.1133,
+      "step": 723
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.230545510604126,
+      "learning_rate": 1.730401529636711e-05,
+      "loss": 1.0752,
+      "step": 724
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.80409338312195,
+      "learning_rate": 1.732791586998088e-05,
+      "loss": 0.749,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.6603722488332275,
+      "learning_rate": 1.7351816443594647e-05,
+      "loss": 1.1504,
+      "step": 726
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.747464754146886,
+      "learning_rate": 1.7375717017208416e-05,
+      "loss": 1.6382,
+      "step": 727
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.420419995486604,
+      "learning_rate": 1.7399617590822183e-05,
+      "loss": 1.5947,
+      "step": 728
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8441085027967685,
+      "learning_rate": 1.742351816443595e-05,
+      "loss": 1.584,
+      "step": 729
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.145445353438485,
+      "learning_rate": 1.7447418738049715e-05,
+      "loss": 0.7966,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.738095714251806,
+      "learning_rate": 1.747131931166348e-05,
+      "loss": 1.1465,
+      "step": 731
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.107596760567716,
+      "learning_rate": 1.7495219885277247e-05,
+      "loss": 1.0996,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.170403181903602,
+      "learning_rate": 1.7519120458891013e-05,
+      "loss": 1.1914,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.189691471344807,
+      "learning_rate": 1.754302103250478e-05,
+      "loss": 1.209,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.533538030229062,
+      "learning_rate": 1.7566921606118546e-05,
+      "loss": 0.9736,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.3694969809845645,
+      "learning_rate": 1.7590822179732312e-05,
+      "loss": 0.9326,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.671807161010067,
+      "learning_rate": 1.761472275334608e-05,
+      "loss": 1.9004,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.244313767674162,
+      "learning_rate": 1.7638623326959848e-05,
+      "loss": 1.0938,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.857719204160044,
+      "learning_rate": 1.7662523900573617e-05,
+      "loss": 1.4863,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.127286834433003,
+      "learning_rate": 1.7686424474187383e-05,
+      "loss": 0.9551,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.141522695289222,
+      "learning_rate": 1.771032504780115e-05,
+      "loss": 0.7734,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4489808221926994,
+      "learning_rate": 1.7734225621414916e-05,
+      "loss": 1.1973,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.633131686776699,
+      "learning_rate": 1.7758126195028682e-05,
+      "loss": 0.9912,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.384185040223198,
+      "learning_rate": 1.7782026768642448e-05,
+      "loss": 1.1855,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5963080548482234,
+      "learning_rate": 1.7805927342256214e-05,
+      "loss": 1.3877,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8514145519014367,
+      "learning_rate": 1.782982791586998e-05,
+      "loss": 1.0479,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.557071498389275,
+      "learning_rate": 1.7853728489483747e-05,
+      "loss": 1.4023,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.780317171514565,
+      "learning_rate": 1.7877629063097516e-05,
+      "loss": 1.6309,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.183421153269972,
+      "learning_rate": 1.7901529636711282e-05,
+      "loss": 1.1221,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.5082954877794235,
+      "learning_rate": 1.792543021032505e-05,
+      "loss": 1.5371,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.227172633483074,
+      "learning_rate": 1.7949330783938815e-05,
+      "loss": 1.2656,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.531836951815115,
+      "learning_rate": 1.7973231357552584e-05,
+      "loss": 1.2168,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.415323394107554,
+      "learning_rate": 1.799713193116635e-05,
+      "loss": 0.7383,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.743656741546147,
+      "learning_rate": 1.8021032504780117e-05,
+      "loss": 1.0146,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.317193675141556,
+      "learning_rate": 1.8044933078393883e-05,
+      "loss": 1.3008,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3779596138273953,
+      "learning_rate": 1.806883365200765e-05,
+      "loss": 1.4473,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.870730601502034,
+      "learning_rate": 1.8092734225621415e-05,
+      "loss": 0.8516,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.597504418569086,
+      "learning_rate": 1.811663479923518e-05,
+      "loss": 1.127,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.167582422161106,
+      "learning_rate": 1.8140535372848947e-05,
+      "loss": 1.293,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.775975128786386,
+      "learning_rate": 1.8164435946462717e-05,
+      "loss": 1.1094,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.719613752309456,
+      "learning_rate": 1.8188336520076483e-05,
+      "loss": 0.7031,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5324074527741773,
+      "learning_rate": 1.821223709369025e-05,
+      "loss": 0.8936,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.075212278863256,
+      "learning_rate": 1.8236137667304016e-05,
+      "loss": 0.7793,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.270175337590091,
+      "learning_rate": 1.826003824091778e-05,
+      "loss": 1.5,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4223739747946964,
+      "learning_rate": 1.828393881453155e-05,
+      "loss": 1.3008,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.48983508143665,
+      "learning_rate": 1.8307839388145317e-05,
+      "loss": 0.8403,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.068342028437575,
+      "learning_rate": 1.8331739961759084e-05,
+      "loss": 1.1777,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.827412825120167,
+      "learning_rate": 1.835564053537285e-05,
+      "loss": 1.1206,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.36622174935057,
+      "learning_rate": 1.8379541108986616e-05,
+      "loss": 1.2617,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5446453060156786,
+      "learning_rate": 1.8403441682600382e-05,
+      "loss": 1.5215,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.482588309823536,
+      "learning_rate": 1.842734225621415e-05,
+      "loss": 1.9609,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.76499926336166,
+      "learning_rate": 1.8451242829827918e-05,
+      "loss": 2.0,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6895267929549744,
+      "learning_rate": 1.8475143403441684e-05,
+      "loss": 1.0205,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6288509303086363,
+      "learning_rate": 1.849904397705545e-05,
+      "loss": 1.2539,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.860820782997195,
+      "learning_rate": 1.8522944550669216e-05,
+      "loss": 0.9766,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.554084646330411,
+      "learning_rate": 1.8546845124282983e-05,
+      "loss": 1.1738,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.787555953515827,
+      "learning_rate": 1.857074569789675e-05,
+      "loss": 1.1406,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.715102003401194,
+      "learning_rate": 1.8594646271510518e-05,
+      "loss": 1.7637,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.945124819947505,
+      "learning_rate": 1.8618546845124284e-05,
+      "loss": 1.0576,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7244776365909766,
+      "learning_rate": 1.864244741873805e-05,
+      "loss": 1.2734,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.223447543805112,
+      "learning_rate": 1.8666347992351817e-05,
+      "loss": 1.0371,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.290396155543927,
+      "learning_rate": 1.8690248565965583e-05,
+      "loss": 0.9131,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.547912445931977,
+      "learning_rate": 1.8714149139579352e-05,
+      "loss": 0.7622,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.844417599847671,
+      "learning_rate": 1.873804971319312e-05,
+      "loss": 1.1895,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.482658269696417,
+      "learning_rate": 1.8761950286806885e-05,
+      "loss": 1.0537,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7429858413334736,
+      "learning_rate": 1.878585086042065e-05,
+      "loss": 1.2559,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.153418671039977,
+      "learning_rate": 1.8809751434034417e-05,
+      "loss": 1.6055,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4024072371821386,
+      "learning_rate": 1.8833652007648183e-05,
+      "loss": 1.042,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4450217057251584,
+      "learning_rate": 1.885755258126195e-05,
+      "loss": 1.2031,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2942549432464263,
+      "learning_rate": 1.8881453154875716e-05,
+      "loss": 1.2012,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.830599916453039,
+      "learning_rate": 1.8905353728489482e-05,
+      "loss": 1.5586,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.275585116467684,
+      "learning_rate": 1.892925430210325e-05,
+      "loss": 0.4712,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.47766795922278,
+      "learning_rate": 1.8953154875717018e-05,
+      "loss": 1.1465,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.64342652634822,
+      "learning_rate": 1.8977055449330787e-05,
+      "loss": 1.25,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 22.14280682395456,
+      "learning_rate": 1.9000956022944553e-05,
+      "loss": 1.3711,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.028305752814597,
+      "learning_rate": 1.902485659655832e-05,
+      "loss": 1.1133,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.8227822203517565,
+      "learning_rate": 1.9048757170172086e-05,
+      "loss": 1.002,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.31350879782802,
+      "learning_rate": 1.9072657743785852e-05,
+      "loss": 0.75,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.567039943188025,
+      "learning_rate": 1.9096558317399618e-05,
+      "loss": 1.2402,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.171357005014779,
+      "learning_rate": 1.9120458891013384e-05,
+      "loss": 1.2012,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.231634963764329,
+      "learning_rate": 1.914435946462715e-05,
+      "loss": 1.3164,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.128635646951533,
+      "learning_rate": 1.9168260038240917e-05,
+      "loss": 1.0811,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.881275509859492,
+      "learning_rate": 1.9192160611854683e-05,
+      "loss": 1.8379,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.126591689709857,
+      "learning_rate": 1.9216061185468452e-05,
+      "loss": 1.1914,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.545110053330103,
+      "learning_rate": 1.923996175908222e-05,
+      "loss": 0.5195,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.010704981226237,
+      "learning_rate": 1.9263862332695988e-05,
+      "loss": 0.8857,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.136589224676035,
+      "learning_rate": 1.9287762906309754e-05,
+      "loss": 1.7168,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.929720967658284,
+      "learning_rate": 1.931166347992352e-05,
+      "loss": 0.8213,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.5368168307763606,
+      "learning_rate": 1.9335564053537286e-05,
+      "loss": 1.0352,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.344938615009774,
+      "learning_rate": 1.9359464627151053e-05,
+      "loss": 1.4316,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.222400706014016,
+      "learning_rate": 1.938336520076482e-05,
+      "loss": 1.6035,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.872098882697245,
+      "learning_rate": 1.9407265774378585e-05,
+      "loss": 1.0107,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.120003253195307,
+      "learning_rate": 1.943116634799235e-05,
+      "loss": 1.6172,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1002678847700285,
+      "learning_rate": 1.9455066921606117e-05,
+      "loss": 1.1191,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.846924953338549,
+      "learning_rate": 1.9478967495219887e-05,
+      "loss": 1.4043,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.442226415530957,
+      "learning_rate": 1.9502868068833653e-05,
+      "loss": 0.9209,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.08223373432311,
+      "learning_rate": 1.952676864244742e-05,
+      "loss": 1.2695,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.875649803842453,
+      "learning_rate": 1.955066921606119e-05,
+      "loss": 1.0264,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.691658616420375,
+      "learning_rate": 1.9574569789674955e-05,
+      "loss": 1.0928,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3750846287042235,
+      "learning_rate": 1.959847036328872e-05,
+      "loss": 1.4805,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.260925708232154,
+      "learning_rate": 1.9622370936902487e-05,
+      "loss": 1.416,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.892657797639856,
+      "learning_rate": 1.9646271510516253e-05,
+      "loss": 1.3555,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.279707123261783,
+      "learning_rate": 1.967017208413002e-05,
+      "loss": 1.1465,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3135622347234115,
+      "learning_rate": 1.9694072657743786e-05,
+      "loss": 1.2285,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.736699519280378,
+      "learning_rate": 1.9717973231357552e-05,
+      "loss": 1.2041,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.722341274928301,
+      "learning_rate": 1.9741873804971318e-05,
+      "loss": 1.5801,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.486933559613196,
+      "learning_rate": 1.9765774378585088e-05,
+      "loss": 1.4443,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.53120084745244,
+      "learning_rate": 1.9789674952198854e-05,
+      "loss": 0.9355,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.292354244719115,
+      "learning_rate": 1.981357552581262e-05,
+      "loss": 1.5352,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.40341316103618,
+      "learning_rate": 1.9837476099426386e-05,
+      "loss": 0.8486,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.35990266908856,
+      "learning_rate": 1.9861376673040156e-05,
+      "loss": 1.2988,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7695410638159164,
+      "learning_rate": 1.9885277246653922e-05,
+      "loss": 0.7158,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.829548434375553,
+      "learning_rate": 1.9909177820267688e-05,
+      "loss": 0.7114,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.404620146464795,
+      "learning_rate": 1.9933078393881454e-05,
+      "loss": 1.0566,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7887278615518065,
+      "learning_rate": 1.995697896749522e-05,
+      "loss": 1.2051,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0760708068714413,
+      "learning_rate": 1.9980879541108987e-05,
+      "loss": 0.4497,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.034400359531306,
+      "learning_rate": 2.0004780114722753e-05,
+      "loss": 0.8706,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.628256367451725,
+      "learning_rate": 2.0028680688336522e-05,
+      "loss": 1.2314,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.380528764009266,
+      "learning_rate": 2.005258126195029e-05,
+      "loss": 1.0996,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3297485220763514,
+      "learning_rate": 2.0076481835564055e-05,
+      "loss": 1.158,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.18410010601667,
+      "learning_rate": 2.010038240917782e-05,
+      "loss": 1.793,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.256286410020168,
+      "learning_rate": 2.0124282982791587e-05,
+      "loss": 1.5547,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5355082530544895,
+      "learning_rate": 2.0148183556405353e-05,
+      "loss": 1.1553,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.382891980225997,
+      "learning_rate": 2.0172084130019123e-05,
+      "loss": 1.8164,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6085760380008196,
+      "learning_rate": 2.019598470363289e-05,
+      "loss": 1.0674,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.860727297202782,
+      "learning_rate": 2.0219885277246655e-05,
+      "loss": 1.1895,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0561349858834634,
+      "learning_rate": 2.024378585086042e-05,
+      "loss": 0.9775,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.307030346595452,
+      "learning_rate": 2.0267686424474187e-05,
+      "loss": 0.832,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.353156271058676,
+      "learning_rate": 2.0291586998087954e-05,
+      "loss": 1.1064,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.0939795266381,
+      "learning_rate": 2.0315487571701723e-05,
+      "loss": 0.7061,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7911328237071045,
+      "learning_rate": 2.033938814531549e-05,
+      "loss": 0.7549,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9090917397754263,
+      "learning_rate": 2.0363288718929256e-05,
+      "loss": 1.0713,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.657880240313784,
+      "learning_rate": 2.0387189292543022e-05,
+      "loss": 0.8447,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6223158055468825,
+      "learning_rate": 2.0411089866156788e-05,
+      "loss": 1.124,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.356396038680081,
+      "learning_rate": 2.0434990439770554e-05,
+      "loss": 0.9209,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.577759460952762,
+      "learning_rate": 2.045889101338432e-05,
+      "loss": 1.041,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8841640539928095,
+      "learning_rate": 2.048279158699809e-05,
+      "loss": 1.334,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.542422625576707,
+      "learning_rate": 2.0506692160611856e-05,
+      "loss": 1.1416,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.008640072146775,
+      "learning_rate": 2.0530592734225622e-05,
+      "loss": 1.623,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6837500203260998,
+      "learning_rate": 2.055449330783939e-05,
+      "loss": 1.0039,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.577142855069086,
+      "learning_rate": 2.0578393881453158e-05,
+      "loss": 1.2109,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.17893713906121,
+      "learning_rate": 2.0602294455066924e-05,
+      "loss": 1.3711,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.095575402747999,
+      "learning_rate": 2.062619502868069e-05,
+      "loss": 1.3945,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.615831829628793,
+      "learning_rate": 2.0650095602294456e-05,
+      "loss": 1.0269,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.164839873990338,
+      "learning_rate": 2.0673996175908223e-05,
+      "loss": 0.9766,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.126525365683329,
+      "learning_rate": 2.069789674952199e-05,
+      "loss": 1.623,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.191818425751879,
+      "learning_rate": 2.0721797323135755e-05,
+      "loss": 1.3145,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.294862978335756,
+      "learning_rate": 2.074569789674952e-05,
+      "loss": 0.8184,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.596816602239309,
+      "learning_rate": 2.0769598470363287e-05,
+      "loss": 1.4629,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.693920440356041,
+      "learning_rate": 2.0793499043977057e-05,
+      "loss": 1.8008,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.725647186683303,
+      "learning_rate": 2.0817399617590823e-05,
+      "loss": 1.0752,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1861029162454426,
+      "learning_rate": 2.084130019120459e-05,
+      "loss": 1.5098,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.54978546380939,
+      "learning_rate": 2.086520076481836e-05,
+      "loss": 1.4258,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.788312903902074,
+      "learning_rate": 2.0889101338432125e-05,
+      "loss": 1.5039,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2236215706218703,
+      "learning_rate": 2.091300191204589e-05,
+      "loss": 1.582,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.565732545259722,
+      "learning_rate": 2.0936902485659657e-05,
+      "loss": 1.1621,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.151076130246526,
+      "learning_rate": 2.0960803059273423e-05,
+      "loss": 1.1426,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.471871972962578,
+      "learning_rate": 2.098470363288719e-05,
+      "loss": 0.9717,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.110988981706816,
+      "learning_rate": 2.1008604206500956e-05,
+      "loss": 1.0596,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.955173285030696,
+      "learning_rate": 2.1032504780114722e-05,
+      "loss": 0.8516,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9762334366377163,
+      "learning_rate": 2.1056405353728488e-05,
+      "loss": 0.9736,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.606003658722111,
+      "learning_rate": 2.1080305927342258e-05,
+      "loss": 0.9316,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.495134241148323,
+      "learning_rate": 2.1104206500956024e-05,
+      "loss": 0.9307,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.476132570296063,
+      "learning_rate": 2.1128107074569793e-05,
+      "loss": 1.2305,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0749307451038455,
+      "learning_rate": 2.115200764818356e-05,
+      "loss": 1.043,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.471794105293937,
+      "learning_rate": 2.1175908221797326e-05,
+      "loss": 1.5469,
+      "step": 886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.867839520815431,
+      "learning_rate": 2.1199808795411092e-05,
+      "loss": 0.9199,
+      "step": 887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.914586441393409,
+      "learning_rate": 2.1223709369024858e-05,
+      "loss": 1.1318,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.291149631838806,
+      "learning_rate": 2.1247609942638624e-05,
+      "loss": 1.0469,
+      "step": 889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0804984145220007,
+      "learning_rate": 2.127151051625239e-05,
+      "loss": 0.9395,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8159248103045673,
+      "learning_rate": 2.1295411089866157e-05,
+      "loss": 0.7002,
+      "step": 891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.368398265516569,
+      "learning_rate": 2.1319311663479923e-05,
+      "loss": 1.1973,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.430323776707594,
+      "learning_rate": 2.134321223709369e-05,
+      "loss": 1.4453,
+      "step": 893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.922479736669258,
+      "learning_rate": 2.136711281070746e-05,
+      "loss": 1.8721,
+      "step": 894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.897363671542192,
+      "learning_rate": 2.1391013384321225e-05,
+      "loss": 0.8477,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.029031454212103,
+      "learning_rate": 2.1414913957934994e-05,
+      "loss": 0.772,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7536375840789242,
+      "learning_rate": 2.143881453154876e-05,
+      "loss": 1.0,
+      "step": 897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.912252587966078,
+      "learning_rate": 2.1462715105162527e-05,
+      "loss": 1.21,
+      "step": 898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.454503625651568,
+      "learning_rate": 2.1486615678776293e-05,
+      "loss": 1.334,
+      "step": 899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.289854801861733,
+      "learning_rate": 2.151051625239006e-05,
+      "loss": 1.623,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7055650872662858,
+      "learning_rate": 2.1534416826003825e-05,
+      "loss": 1.0996,
+      "step": 901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8397386366164037,
+      "learning_rate": 2.155831739961759e-05,
+      "loss": 1.3242,
+      "step": 902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.646401249453479,
+      "learning_rate": 2.1582217973231357e-05,
+      "loss": 1.6992,
+      "step": 903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.56372557857499,
+      "learning_rate": 2.1606118546845124e-05,
+      "loss": 0.9932,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6896115369275866,
+      "learning_rate": 2.1630019120458893e-05,
+      "loss": 1.0449,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2949334524841056,
+      "learning_rate": 2.165391969407266e-05,
+      "loss": 1.4297,
+      "step": 906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8772984732836515,
+      "learning_rate": 2.1677820267686425e-05,
+      "loss": 1.1284,
+      "step": 907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.278095939042715,
+      "learning_rate": 2.170172084130019e-05,
+      "loss": 1.1057,
+      "step": 908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.118481513260478,
+      "learning_rate": 2.172562141491396e-05,
+      "loss": 1.1016,
+      "step": 909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7084095264851062,
+      "learning_rate": 2.1749521988527727e-05,
+      "loss": 1.1289,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.446406122068163,
+      "learning_rate": 2.1773422562141494e-05,
+      "loss": 1.4355,
+      "step": 911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.07460229191059,
+      "learning_rate": 2.179732313575526e-05,
+      "loss": 1.5098,
+      "step": 912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7142011245066056,
+      "learning_rate": 2.1821223709369026e-05,
+      "loss": 1.21,
+      "step": 913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.336495688310589,
+      "learning_rate": 2.1845124282982792e-05,
+      "loss": 0.8232,
+      "step": 914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.8302013290493795,
+      "learning_rate": 2.1869024856596558e-05,
+      "loss": 0.7378,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7221059322379286,
+      "learning_rate": 2.1892925430210324e-05,
+      "loss": 1.3281,
+      "step": 916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.697003323548899,
+      "learning_rate": 2.1916826003824094e-05,
+      "loss": 1.196,
+      "step": 917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.717507606788818,
+      "learning_rate": 2.194072657743786e-05,
+      "loss": 1.3047,
+      "step": 918
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.957211366758891,
+      "learning_rate": 2.1964627151051626e-05,
+      "loss": 1.2969,
+      "step": 919
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.4515487010340555,
+      "learning_rate": 2.1988527724665392e-05,
+      "loss": 1.1484,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.674737381425199,
+      "learning_rate": 2.201242829827916e-05,
+      "loss": 1.2969,
+      "step": 921
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.916138421253892,
+      "learning_rate": 2.2036328871892928e-05,
+      "loss": 1.5127,
+      "step": 922
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.8391055574801785,
+      "learning_rate": 2.2060229445506694e-05,
+      "loss": 1.0674,
+      "step": 923
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.419656257554033,
+      "learning_rate": 2.208413001912046e-05,
+      "loss": 0.6201,
+      "step": 924
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7992947877730434,
+      "learning_rate": 2.2108030592734227e-05,
+      "loss": 1.2437,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9077536682048044,
+      "learning_rate": 2.2131931166347993e-05,
+      "loss": 1.0723,
+      "step": 926
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.181536501729523,
+      "learning_rate": 2.215583173996176e-05,
+      "loss": 1.3867,
+      "step": 927
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.092078560725818,
+      "learning_rate": 2.217973231357553e-05,
+      "loss": 0.9346,
+      "step": 928
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.477393420604551,
+      "learning_rate": 2.2203632887189295e-05,
+      "loss": 1.0225,
+      "step": 929
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.15598479866473,
+      "learning_rate": 2.222753346080306e-05,
+      "loss": 1.6768,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.2977859478943685,
+      "learning_rate": 2.2251434034416827e-05,
+      "loss": 1.3926,
+      "step": 931
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.045313471281628,
+      "learning_rate": 2.2275334608030593e-05,
+      "loss": 1.3555,
+      "step": 932
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.263695836050582,
+      "learning_rate": 2.229923518164436e-05,
+      "loss": 1.1709,
+      "step": 933
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1439440445519122,
+      "learning_rate": 2.2323135755258126e-05,
+      "loss": 0.874,
+      "step": 934
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.877457250480912,
+      "learning_rate": 2.2347036328871895e-05,
+      "loss": 1.1016,
+      "step": 935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9746610820884163,
+      "learning_rate": 2.237093690248566e-05,
+      "loss": 1.5391,
+      "step": 936
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9902107741148973,
+      "learning_rate": 2.2394837476099428e-05,
+      "loss": 0.501,
+      "step": 937
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6929199990253863,
+      "learning_rate": 2.2418738049713194e-05,
+      "loss": 0.9424,
+      "step": 938
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.913955291071656,
+      "learning_rate": 2.244263862332696e-05,
+      "loss": 1.2158,
+      "step": 939
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.414232530948078,
+      "learning_rate": 2.246653919694073e-05,
+      "loss": 1.375,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.030386774267884,
+      "learning_rate": 2.2490439770554496e-05,
+      "loss": 1.4199,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.193595294228796,
+      "learning_rate": 2.2514340344168262e-05,
+      "loss": 1.3252,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.165108514887087,
+      "learning_rate": 2.2538240917782028e-05,
+      "loss": 0.7295,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.409398733592484,
+      "learning_rate": 2.2562141491395794e-05,
+      "loss": 1.0742,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.220364804423837,
+      "learning_rate": 2.258604206500956e-05,
+      "loss": 0.9551,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.311533938975675,
+      "learning_rate": 2.2609942638623326e-05,
+      "loss": 1.2002,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.50308782423071,
+      "learning_rate": 2.2633843212237093e-05,
+      "loss": 0.998,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5483084987370352,
+      "learning_rate": 2.2657743785850862e-05,
+      "loss": 1.5117,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.461533471454382,
+      "learning_rate": 2.268164435946463e-05,
+      "loss": 0.5217,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5881617450846233,
+      "learning_rate": 2.2705544933078395e-05,
+      "loss": 0.6578,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.971965786352687,
+      "learning_rate": 2.2729445506692164e-05,
+      "loss": 2.0977,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.471864556435055,
+      "learning_rate": 2.275334608030593e-05,
+      "loss": 1.2266,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.071185408288612,
+      "learning_rate": 2.2777246653919696e-05,
+      "loss": 1.1113,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.793682614167431,
+      "learning_rate": 2.2801147227533463e-05,
+      "loss": 1.3262,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.81197361507931,
+      "learning_rate": 2.282504780114723e-05,
+      "loss": 0.8506,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9302258619933927,
+      "learning_rate": 2.2848948374760995e-05,
+      "loss": 1.1045,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.270728787109471,
+      "learning_rate": 2.287284894837476e-05,
+      "loss": 1.498,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7917461001681385,
+      "learning_rate": 2.2896749521988527e-05,
+      "loss": 0.73,
+      "step": 958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.26288749176792,
+      "learning_rate": 2.2920650095602293e-05,
+      "loss": 1.6641,
+      "step": 959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.252170623982773,
+      "learning_rate": 2.294455066921606e-05,
+      "loss": 1.3057,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3427342566218767,
+      "learning_rate": 2.296845124282983e-05,
+      "loss": 0.5854,
+      "step": 961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2935406786657984,
+      "learning_rate": 2.2992351816443595e-05,
+      "loss": 0.917,
+      "step": 962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.002424697310569,
+      "learning_rate": 2.3016252390057365e-05,
+      "loss": 0.7002,
+      "step": 963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.855108123423897,
+      "learning_rate": 2.304015296367113e-05,
+      "loss": 0.9199,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.327581158014165,
+      "learning_rate": 2.3064053537284897e-05,
+      "loss": 1.0176,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.331021503234098,
+      "learning_rate": 2.3087954110898663e-05,
+      "loss": 1.4355,
+      "step": 966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.48166266962616,
+      "learning_rate": 2.311185468451243e-05,
+      "loss": 0.959,
+      "step": 967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.966035075653325,
+      "learning_rate": 2.3135755258126196e-05,
+      "loss": 0.9521,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.92595388860145,
+      "learning_rate": 2.3159655831739962e-05,
+      "loss": 2.0781,
+      "step": 969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.827335253022864,
+      "learning_rate": 2.3183556405353728e-05,
+      "loss": 0.6494,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.952878291106788,
+      "learning_rate": 2.3207456978967494e-05,
+      "loss": 1.5586,
+      "step": 971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.702289095993876,
+      "learning_rate": 2.323135755258126e-05,
+      "loss": 1.3975,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.055105663663681,
+      "learning_rate": 2.325525812619503e-05,
+      "loss": 1.1943,
+      "step": 973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8073565698911063,
+      "learning_rate": 2.3279158699808796e-05,
+      "loss": 0.8701,
+      "step": 974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.324325711823576,
+      "learning_rate": 2.3303059273422566e-05,
+      "loss": 1.5684,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9224392306481235,
+      "learning_rate": 2.3326959847036332e-05,
+      "loss": 1.1289,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.751099293176778,
+      "learning_rate": 2.3350860420650098e-05,
+      "loss": 1.8477,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.045625589513007,
+      "learning_rate": 2.3374760994263864e-05,
+      "loss": 1.3848,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8113205025916623,
+      "learning_rate": 2.339866156787763e-05,
+      "loss": 1.1582,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.203216253702044,
+      "learning_rate": 2.3422562141491397e-05,
+      "loss": 1.7559,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4030915061626548,
+      "learning_rate": 2.3446462715105163e-05,
+      "loss": 0.8301,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.712160068869524,
+      "learning_rate": 2.347036328871893e-05,
+      "loss": 1.127,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.189697418046684,
+      "learning_rate": 2.3494263862332695e-05,
+      "loss": 0.9111,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.902571112491699,
+      "learning_rate": 2.3518164435946465e-05,
+      "loss": 1.1035,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.498038423443866,
+      "learning_rate": 2.354206500956023e-05,
+      "loss": 1.2588,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8872462542476143,
+      "learning_rate": 2.3565965583173997e-05,
+      "loss": 1.3301,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.188054504033913,
+      "learning_rate": 2.3589866156787763e-05,
+      "loss": 1.3848,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.581358795769103,
+      "learning_rate": 2.3613766730401533e-05,
+      "loss": 0.9707,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3142531452082022,
+      "learning_rate": 2.36376673040153e-05,
+      "loss": 1.0742,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.59730680232386,
+      "learning_rate": 2.3661567877629065e-05,
+      "loss": 1.4629,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.637481443511625,
+      "learning_rate": 2.368546845124283e-05,
+      "loss": 0.9907,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.004120845198446,
+      "learning_rate": 2.3709369024856597e-05,
+      "loss": 0.6338,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.076612793170485,
+      "learning_rate": 2.3733269598470364e-05,
+      "loss": 1.4883,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5489915086908512,
+      "learning_rate": 2.375717017208413e-05,
+      "loss": 1.3359,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.50067451508085,
+      "learning_rate": 2.3781070745697896e-05,
+      "loss": 0.9287,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.921076536666455,
+      "learning_rate": 2.3804971319311666e-05,
+      "loss": 2.3135,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.974455209502468,
+      "learning_rate": 2.3828871892925432e-05,
+      "loss": 1.5195,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0973422070476153,
+      "learning_rate": 2.3852772466539198e-05,
+      "loss": 1.1182,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5601172258600187,
+      "learning_rate": 2.3876673040152964e-05,
+      "loss": 1.248,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.403695089499746,
+      "learning_rate": 2.390057361376673e-05,
+      "loss": 0.9961,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.672254668153485,
+      "learning_rate": 2.39244741873805e-05,
+      "loss": 0.4756,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.480545163022659,
+      "learning_rate": 2.3948374760994266e-05,
+      "loss": 1.25,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.163883072190968,
+      "learning_rate": 2.3972275334608032e-05,
+      "loss": 1.5312,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.983513772580493,
+      "learning_rate": 2.3996175908221798e-05,
+      "loss": 1.3257,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.094000940185648,
+      "learning_rate": 2.4020076481835564e-05,
+      "loss": 1.916,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.248182834520823,
+      "learning_rate": 2.404397705544933e-05,
+      "loss": 1.5977,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.953986635655387,
+      "learning_rate": 2.40678776290631e-05,
+      "loss": 0.8271,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.216840200412098,
+      "learning_rate": 2.4091778202676866e-05,
+      "loss": 1.2578,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.722812545878341,
+      "learning_rate": 2.4115678776290633e-05,
+      "loss": 0.9404,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.075712015833167,
+      "learning_rate": 2.41395793499044e-05,
+      "loss": 0.9941,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.439113446369147,
+      "learning_rate": 2.4163479923518165e-05,
+      "loss": 1.3164,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.018673700584259,
+      "learning_rate": 2.418738049713193e-05,
+      "loss": 1.1514,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.7455802184913525,
+      "learning_rate": 2.4211281070745697e-05,
+      "loss": 1.582,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.768716578535202,
+      "learning_rate": 2.4235181644359467e-05,
+      "loss": 0.8857,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.965287447596147,
+      "learning_rate": 2.4259082217973233e-05,
+      "loss": 0.8604,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.112365794673331,
+      "learning_rate": 2.4282982791587e-05,
+      "loss": 1.0762,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.261325177415125,
+      "learning_rate": 2.4306883365200765e-05,
+      "loss": 1.2109,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.771931792102676,
+      "learning_rate": 2.433078393881453e-05,
+      "loss": 0.5698,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.490971939818083,
+      "learning_rate": 2.43546845124283e-05,
+      "loss": 0.96,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.63833902248787,
+      "learning_rate": 2.4378585086042067e-05,
+      "loss": 1.332,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.42792744072175,
+      "learning_rate": 2.4402485659655833e-05,
+      "loss": 1.0713,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.743191145967515,
+      "learning_rate": 2.44263862332696e-05,
+      "loss": 1.3125,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.450729850323599,
+      "learning_rate": 2.4450286806883366e-05,
+      "loss": 0.9697,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.347790909394582,
+      "learning_rate": 2.4474187380497132e-05,
+      "loss": 1.1455,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8503328363249145,
+      "learning_rate": 2.4498087954110898e-05,
+      "loss": 1.2119,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2161637072767912,
+      "learning_rate": 2.4521988527724664e-05,
+      "loss": 1.2109,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.5407591724448055,
+      "learning_rate": 2.4545889101338434e-05,
+      "loss": 1.0479,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4254942265914172,
+      "learning_rate": 2.45697896749522e-05,
+      "loss": 1.4727,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.402218382395397,
+      "learning_rate": 2.4593690248565966e-05,
+      "loss": 1.1719,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.763216431298205,
+      "learning_rate": 2.4617590822179736e-05,
+      "loss": 0.9482,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9659425008786107,
+      "learning_rate": 2.4641491395793502e-05,
+      "loss": 0.6821,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2754919483433493,
+      "learning_rate": 2.4665391969407268e-05,
+      "loss": 0.9375,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.81766266800663,
+      "learning_rate": 2.4689292543021034e-05,
+      "loss": 1.1113,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6929309743777443,
+      "learning_rate": 2.47131931166348e-05,
+      "loss": 0.9502,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.507711091459297,
+      "learning_rate": 2.4737093690248567e-05,
+      "loss": 1.3848,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.145023400612157,
+      "learning_rate": 2.4760994263862333e-05,
+      "loss": 0.521,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.194331990122349,
+      "learning_rate": 2.47848948374761e-05,
+      "loss": 1.1201,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9682207130068043,
+      "learning_rate": 2.4808795411089865e-05,
+      "loss": 1.084,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.234258394223579,
+      "learning_rate": 2.483269598470363e-05,
+      "loss": 0.8682,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.45563082335764,
+      "learning_rate": 2.48565965583174e-05,
+      "loss": 0.7651,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.059914496920817,
+      "learning_rate": 2.4880497131931167e-05,
+      "loss": 2.0332,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.418749671946814,
+      "learning_rate": 2.4904397705544937e-05,
+      "loss": 1.332,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.336503605358482,
+      "learning_rate": 2.4928298279158703e-05,
+      "loss": 1.0234,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.774908315157771,
+      "learning_rate": 2.495219885277247e-05,
+      "loss": 0.8423,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.055454694913061,
+      "learning_rate": 2.4976099426386235e-05,
+      "loss": 1.1328,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.018263993491921,
+      "learning_rate": 2.5e-05,
+      "loss": 0.835,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.811316208338027,
+      "learning_rate": 2.5023900573613767e-05,
+      "loss": 1.2168,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.470079557662375,
+      "learning_rate": 2.5047801147227534e-05,
+      "loss": 1.4863,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.621020876872356,
+      "learning_rate": 2.50717017208413e-05,
+      "loss": 1.3965,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.077250926078886,
+      "learning_rate": 2.5095602294455066e-05,
+      "loss": 1.4717,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.184314588330783,
+      "learning_rate": 2.5119502868068835e-05,
+      "loss": 0.917,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.542321111522018,
+      "learning_rate": 2.51434034416826e-05,
+      "loss": 1.1562,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.891144688177127,
+      "learning_rate": 2.5167304015296368e-05,
+      "loss": 1.0996,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.066563611906604,
+      "learning_rate": 2.5191204588910134e-05,
+      "loss": 0.75,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.888825552400939,
+      "learning_rate": 2.52151051625239e-05,
+      "loss": 0.98,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7854627361164024,
+      "learning_rate": 2.5239005736137666e-05,
+      "loss": 0.8203,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.814209990419731,
+      "learning_rate": 2.5262906309751432e-05,
+      "loss": 1.1953,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1150408655411,
+      "learning_rate": 2.52868068833652e-05,
+      "loss": 1.4785,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.823764434077702,
+      "learning_rate": 2.5310707456978965e-05,
+      "loss": 1.1914,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.405979081223368,
+      "learning_rate": 2.5334608030592738e-05,
+      "loss": 1.2969,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.444146597221717,
+      "learning_rate": 2.5358508604206504e-05,
+      "loss": 1.625,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.013227389197707,
+      "learning_rate": 2.538240917782027e-05,
+      "loss": 1.1992,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9065064347867087,
+      "learning_rate": 2.5406309751434036e-05,
+      "loss": 0.8818,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.165368173333714,
+      "learning_rate": 2.5430210325047802e-05,
+      "loss": 0.5618,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.194728022091693,
+      "learning_rate": 2.5454110898661572e-05,
+      "loss": 1.2676,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.761291334514109,
+      "learning_rate": 2.5478011472275338e-05,
+      "loss": 1.2734,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.865294552204616,
+      "learning_rate": 2.5501912045889104e-05,
+      "loss": 0.9419,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3978957171859476,
+      "learning_rate": 2.552581261950287e-05,
+      "loss": 1.1328,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.987718980664175,
+      "learning_rate": 2.5549713193116637e-05,
+      "loss": 1.3008,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.571676641235047,
+      "learning_rate": 2.5573613766730403e-05,
+      "loss": 0.9946,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.935826384886688,
+      "learning_rate": 2.559751434034417e-05,
+      "loss": 1.25,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.896262724964317,
+      "learning_rate": 2.5621414913957935e-05,
+      "loss": 1.3984,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.466689219222446,
+      "learning_rate": 2.56453154875717e-05,
+      "loss": 1.1211,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.026241434441913,
+      "learning_rate": 2.566921606118547e-05,
+      "loss": 1.4121,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.233703248666991,
+      "learning_rate": 2.5693116634799237e-05,
+      "loss": 0.9414,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.364775104419094,
+      "learning_rate": 2.5717017208413003e-05,
+      "loss": 1.0586,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.073067404582864,
+      "learning_rate": 2.574091778202677e-05,
+      "loss": 0.8936,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.585042815151045,
+      "learning_rate": 2.5764818355640536e-05,
+      "loss": 0.7725,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.620057201976792,
+      "learning_rate": 2.5788718929254302e-05,
+      "loss": 0.918,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.43744710269877,
+      "learning_rate": 2.5812619502868068e-05,
+      "loss": 1.0439,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8483008556474907,
+      "learning_rate": 2.5836520076481834e-05,
+      "loss": 1.1331,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.606883906458465,
+      "learning_rate": 2.58604206500956e-05,
+      "loss": 1.1738,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8984653029432184,
+      "learning_rate": 2.5884321223709366e-05,
+      "loss": 1.541,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8294155485172916,
+      "learning_rate": 2.5908221797323136e-05,
+      "loss": 1.2871,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.887747807592706,
+      "learning_rate": 2.5932122370936902e-05,
+      "loss": 1.6289,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4885095263177885,
+      "learning_rate": 2.5956022944550672e-05,
+      "loss": 0.9751,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.959813999561502,
+      "learning_rate": 2.5979923518164438e-05,
+      "loss": 1.2656,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.994050254898765,
+      "learning_rate": 2.6003824091778207e-05,
+      "loss": 1.1484,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.108785089634749,
+      "learning_rate": 2.6027724665391974e-05,
+      "loss": 1.0459,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.147074256231259,
+      "learning_rate": 2.605162523900574e-05,
+      "loss": 1.5371,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.891345807774968,
+      "learning_rate": 2.6075525812619506e-05,
+      "loss": 1.4688,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.933198997319816,
+      "learning_rate": 2.6099426386233272e-05,
+      "loss": 1.042,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.242711262763562,
+      "learning_rate": 2.612332695984704e-05,
+      "loss": 1.3291,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9435203293899328,
+      "learning_rate": 2.6147227533460805e-05,
+      "loss": 0.5017,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.574426597013362,
+      "learning_rate": 2.617112810707457e-05,
+      "loss": 0.5693,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.656383805394154,
+      "learning_rate": 2.6195028680688337e-05,
+      "loss": 1.1777,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.22131826716717,
+      "learning_rate": 2.6218929254302106e-05,
+      "loss": 2.3008,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.146316032709827,
+      "learning_rate": 2.6242829827915873e-05,
+      "loss": 1.7148,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.720744843756741,
+      "learning_rate": 2.626673040152964e-05,
+      "loss": 1.5312,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.2756388904788585,
+      "learning_rate": 2.6290630975143405e-05,
+      "loss": 0.9756,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.530310041520123,
+      "learning_rate": 2.631453154875717e-05,
+      "loss": 1.0906,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.18456052720608,
+      "learning_rate": 2.6338432122370937e-05,
+      "loss": 1.7305,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.028850176803675,
+      "learning_rate": 2.6362332695984703e-05,
+      "loss": 1.373,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.58434421251175,
+      "learning_rate": 2.638623326959847e-05,
+      "loss": 1.0635,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.928042669777474,
+      "learning_rate": 2.6410133843212236e-05,
+      "loss": 0.998,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.741239685550006,
+      "learning_rate": 2.6434034416826002e-05,
+      "loss": 1.0254,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.027909052286898,
+      "learning_rate": 2.645793499043977e-05,
+      "loss": 1.7383,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3460948957449737,
+      "learning_rate": 2.6481835564053538e-05,
+      "loss": 1.4043,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.659987100851483,
+      "learning_rate": 2.6505736137667304e-05,
+      "loss": 1.1245,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.790246305864434,
+      "learning_rate": 2.652963671128107e-05,
+      "loss": 1.5879,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1343129053934637,
+      "learning_rate": 2.6553537284894836e-05,
+      "loss": 0.9941,
+      "step": 1111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.897272004183098,
+      "learning_rate": 2.657743785850861e-05,
+      "loss": 1.2764,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.93593372497741,
+      "learning_rate": 2.6601338432122375e-05,
+      "loss": 1.1445,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.602402968513283,
+      "learning_rate": 2.662523900573614e-05,
+      "loss": 1.5117,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.778160410991238,
+      "learning_rate": 2.6649139579349908e-05,
+      "loss": 0.7461,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.757278756468979,
+      "learning_rate": 2.6673040152963674e-05,
+      "loss": 0.915,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7299918144483977,
+      "learning_rate": 2.669694072657744e-05,
+      "loss": 1.292,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.519574928731348,
+      "learning_rate": 2.6720841300191206e-05,
+      "loss": 1.375,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.247100535099293,
+      "learning_rate": 2.6744741873804972e-05,
+      "loss": 2.084,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.416983493397892,
+      "learning_rate": 2.6768642447418742e-05,
+      "loss": 1.3887,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.934552875188794,
+      "learning_rate": 2.6792543021032508e-05,
+      "loss": 1.7266,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.884195438960564,
+      "learning_rate": 2.6816443594646274e-05,
+      "loss": 0.958,
+      "step": 1122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.035380188863674,
+      "learning_rate": 2.684034416826004e-05,
+      "loss": 1.8633,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6994150627275455,
+      "learning_rate": 2.6864244741873807e-05,
+      "loss": 0.6895,
+      "step": 1124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.096761924663464,
+      "learning_rate": 2.6888145315487573e-05,
+      "loss": 0.7441,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.567810952145208,
+      "learning_rate": 2.691204588910134e-05,
+      "loss": 1.1719,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.90286600050994,
+      "learning_rate": 2.6935946462715105e-05,
+      "loss": 2.1172,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.631777153953095,
+      "learning_rate": 2.695984703632887e-05,
+      "loss": 0.9639,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.248887262041739,
+      "learning_rate": 2.6983747609942637e-05,
+      "loss": 1.0879,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.471811648085935,
+      "learning_rate": 2.7007648183556407e-05,
+      "loss": 1.1328,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.507815294533168,
+      "learning_rate": 2.7031548757170173e-05,
+      "loss": 1.2324,
+      "step": 1131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.665226094561078,
+      "learning_rate": 2.705544933078394e-05,
+      "loss": 1.2578,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.84307563684993,
+      "learning_rate": 2.7079349904397706e-05,
+      "loss": 1.3047,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.6046345884678175,
+      "learning_rate": 2.7103250478011472e-05,
+      "loss": 1.3535,
+      "step": 1134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.005725410952659,
+      "learning_rate": 2.7127151051625238e-05,
+      "loss": 1.9023,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.434459672012247,
+      "learning_rate": 2.7151051625239004e-05,
+      "loss": 1.2451,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.732586110374988,
+      "learning_rate": 2.717495219885277e-05,
+      "loss": 0.8623,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.589064096016066,
+      "learning_rate": 2.7198852772466543e-05,
+      "loss": 1.375,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.028189510186986,
+      "learning_rate": 2.722275334608031e-05,
+      "loss": 0.8262,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.425130878819036,
+      "learning_rate": 2.7246653919694075e-05,
+      "loss": 1.5176,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.266983455740208,
+      "learning_rate": 2.727055449330784e-05,
+      "loss": 2.1875,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.520077851480926,
+      "learning_rate": 2.7294455066921608e-05,
+      "loss": 0.9365,
+      "step": 1142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.831811182905534,
+      "learning_rate": 2.7318355640535377e-05,
+      "loss": 1.0293,
+      "step": 1143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.022105409414139,
+      "learning_rate": 2.7342256214149144e-05,
+      "loss": 0.7202,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.461884581131825,
+      "learning_rate": 2.736615678776291e-05,
+      "loss": 1.3027,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4931730718712446,
+      "learning_rate": 2.7390057361376676e-05,
+      "loss": 1.207,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.24647962202418,
+      "learning_rate": 2.7413957934990442e-05,
+      "loss": 1.2656,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.83975567884637,
+      "learning_rate": 2.7437858508604208e-05,
+      "loss": 0.9375,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.306483356857629,
+      "learning_rate": 2.7461759082217974e-05,
+      "loss": 1.0176,
+      "step": 1149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.857833176130512,
+      "learning_rate": 2.748565965583174e-05,
+      "loss": 1.2285,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9181195772912742,
+      "learning_rate": 2.7509560229445507e-05,
+      "loss": 0.9385,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.899478694494771,
+      "learning_rate": 2.7533460803059273e-05,
+      "loss": 1.3477,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6909588578962214,
+      "learning_rate": 2.7557361376673043e-05,
+      "loss": 1.1426,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5733086433079575,
+      "learning_rate": 2.758126195028681e-05,
+      "loss": 0.6926,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.638417929571338,
+      "learning_rate": 2.7605162523900575e-05,
+      "loss": 1.0898,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7344666792976184,
+      "learning_rate": 2.762906309751434e-05,
+      "loss": 1.8164,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8724700742910714,
+      "learning_rate": 2.7652963671128107e-05,
+      "loss": 1.2422,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.641428808179702,
+      "learning_rate": 2.7676864244741873e-05,
+      "loss": 1.0713,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.816554982230247,
+      "learning_rate": 2.770076481835564e-05,
+      "loss": 1.627,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6041335323038126,
+      "learning_rate": 2.7724665391969406e-05,
+      "loss": 1.2266,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.954158218723229,
+      "learning_rate": 2.7748565965583172e-05,
+      "loss": 1.0781,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7113364557337953,
+      "learning_rate": 2.7772466539196938e-05,
+      "loss": 0.5957,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.836411953304129,
+      "learning_rate": 2.7796367112810708e-05,
+      "loss": 1.1719,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.53642112083214,
+      "learning_rate": 2.7820267686424477e-05,
+      "loss": 0.8584,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.617758778506583,
+      "learning_rate": 2.7844168260038243e-05,
+      "loss": 0.9141,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.0948868869433275,
+      "learning_rate": 2.7868068833652013e-05,
+      "loss": 1.1465,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.107855037131515,
+      "learning_rate": 2.789196940726578e-05,
+      "loss": 1.5859,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.834001120835442,
+      "learning_rate": 2.7915869980879545e-05,
+      "loss": 0.887,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7886631783957494,
+      "learning_rate": 2.793977055449331e-05,
+      "loss": 0.647,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.455126479653898,
+      "learning_rate": 2.7963671128107078e-05,
+      "loss": 1.1348,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3169016833082945,
+      "learning_rate": 2.7987571701720844e-05,
+      "loss": 1.3301,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9552790137950495,
+      "learning_rate": 2.801147227533461e-05,
+      "loss": 0.8379,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6791320758997204,
+      "learning_rate": 2.8035372848948376e-05,
+      "loss": 1.1123,
+      "step": 1173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.29681147273436,
+      "learning_rate": 2.8059273422562142e-05,
+      "loss": 1.2695,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.019909424781517,
+      "learning_rate": 2.808317399617591e-05,
+      "loss": 1.2266,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2413814415928193,
+      "learning_rate": 2.8107074569789678e-05,
+      "loss": 1.2236,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.619752641188157,
+      "learning_rate": 2.8130975143403444e-05,
+      "loss": 1.2324,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2505865392176796,
+      "learning_rate": 2.815487571701721e-05,
+      "loss": 1.375,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7525211760084183,
+      "learning_rate": 2.8178776290630977e-05,
+      "loss": 1.5195,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.994115486974822,
+      "learning_rate": 2.8202676864244743e-05,
+      "loss": 1.1719,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.617352745191452,
+      "learning_rate": 2.822657743785851e-05,
+      "loss": 1.4238,
+      "step": 1181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9507596021937568,
+      "learning_rate": 2.8250478011472275e-05,
+      "loss": 0.8716,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.215363433846748,
+      "learning_rate": 2.827437858508604e-05,
+      "loss": 0.9316,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.659452030664412,
+      "learning_rate": 2.8298279158699807e-05,
+      "loss": 0.7798,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.602320558447127,
+      "learning_rate": 2.8322179732313574e-05,
+      "loss": 1.1758,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6195192795823274,
+      "learning_rate": 2.8346080305927343e-05,
+      "loss": 1.4316,
+      "step": 1186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8910371493152147,
+      "learning_rate": 2.836998087954111e-05,
+      "loss": 1.0508,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.992547291284458,
+      "learning_rate": 2.8393881453154875e-05,
+      "loss": 1.4785,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8717418951127898,
+      "learning_rate": 2.841778202676864e-05,
+      "loss": 1.7012,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.082428435720068,
+      "learning_rate": 2.8441682600382415e-05,
+      "loss": 1.3838,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.32877826880935,
+      "learning_rate": 2.846558317399618e-05,
+      "loss": 1.4707,
+      "step": 1191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.780376325307687,
+      "learning_rate": 2.8489483747609947e-05,
+      "loss": 1.3984,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.686465490024702,
+      "learning_rate": 2.8513384321223713e-05,
+      "loss": 1.0127,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.645022957721258,
+      "learning_rate": 2.853728489483748e-05,
+      "loss": 1.5156,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.497670524379706,
+      "learning_rate": 2.8561185468451245e-05,
+      "loss": 1.3457,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5579214287971075,
+      "learning_rate": 2.858508604206501e-05,
+      "loss": 1.5918,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.443170759776727,
+      "learning_rate": 2.8608986615678778e-05,
+      "loss": 1.0107,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.145785724043845,
+      "learning_rate": 2.8632887189292544e-05,
+      "loss": 0.7949,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.565402427224764,
+      "learning_rate": 2.8656787762906313e-05,
+      "loss": 1.4434,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.624584685568063,
+      "learning_rate": 2.868068833652008e-05,
+      "loss": 1.4629,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.988000453500061,
+      "learning_rate": 2.8704588910133846e-05,
+      "loss": 1.3691,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.565794060325899,
+      "learning_rate": 2.8728489483747612e-05,
+      "loss": 0.897,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.705639933419054,
+      "learning_rate": 2.8752390057361378e-05,
+      "loss": 0.4601,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1377886210060018,
+      "learning_rate": 2.8776290630975144e-05,
+      "loss": 1.2441,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.833350649014088,
+      "learning_rate": 2.880019120458891e-05,
+      "loss": 1.1016,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1369379520426954,
+      "learning_rate": 2.8824091778202677e-05,
+      "loss": 1.0977,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.360100361080168,
+      "learning_rate": 2.8847992351816443e-05,
+      "loss": 1.5996,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.272047523701118,
+      "learning_rate": 2.887189292543021e-05,
+      "loss": 1.1709,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.761757812367586,
+      "learning_rate": 2.889579349904398e-05,
+      "loss": 1.6074,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.521881481337439,
+      "learning_rate": 2.8919694072657745e-05,
+      "loss": 1.373,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.69672151375969,
+      "learning_rate": 2.894359464627151e-05,
+      "loss": 1.2305,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.128196331962907,
+      "learning_rate": 2.8967495219885277e-05,
+      "loss": 1.4023,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.228812596612894,
+      "learning_rate": 2.8991395793499043e-05,
+      "loss": 0.9414,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7696643232663862,
+      "learning_rate": 2.901529636711281e-05,
+      "loss": 1.6152,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8200907527578405,
+      "learning_rate": 2.9039196940726576e-05,
+      "loss": 1.0908,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.3610056184306245,
+      "learning_rate": 2.906309751434035e-05,
+      "loss": 1.1797,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.362000047464728,
+      "learning_rate": 2.9086998087954115e-05,
+      "loss": 1.3369,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.486818302514926,
+      "learning_rate": 2.911089866156788e-05,
+      "loss": 1.6582,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3500500803378275,
+      "learning_rate": 2.9134799235181647e-05,
+      "loss": 1.2456,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9049182300090024,
+      "learning_rate": 2.9158699808795413e-05,
+      "loss": 1.2344,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.796288284905948,
+      "learning_rate": 2.918260038240918e-05,
+      "loss": 1.2085,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.923234266417076,
+      "learning_rate": 2.920650095602295e-05,
+      "loss": 1.1484,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9112733813300467,
+      "learning_rate": 2.9230401529636715e-05,
+      "loss": 0.8525,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.170261117593313,
+      "learning_rate": 2.925430210325048e-05,
+      "loss": 1.5215,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9850656399306907,
+      "learning_rate": 2.9278202676864247e-05,
+      "loss": 1.5586,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.585828629094532,
+      "learning_rate": 2.9302103250478014e-05,
+      "loss": 1.1973,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.131549144538914,
+      "learning_rate": 2.932600382409178e-05,
+      "loss": 0.9746,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6967556415808134,
+      "learning_rate": 2.9349904397705546e-05,
+      "loss": 1.1104,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.40919134806656,
+      "learning_rate": 2.9373804971319312e-05,
+      "loss": 1.6113,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.100575541876248,
+      "learning_rate": 2.939770554493308e-05,
+      "loss": 1.1328,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.94123644852968,
+      "learning_rate": 2.9421606118546845e-05,
+      "loss": 1.416,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.959634282571168,
+      "learning_rate": 2.9445506692160614e-05,
+      "loss": 1.9355,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9459083824543733,
+      "learning_rate": 2.946940726577438e-05,
+      "loss": 0.9746,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.465635809094674,
+      "learning_rate": 2.9493307839388146e-05,
+      "loss": 1.5996,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.930654625526292,
+      "learning_rate": 2.9517208413001913e-05,
+      "loss": 1.083,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.558774773927191,
+      "learning_rate": 2.954110898661568e-05,
+      "loss": 0.6328,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.787554785294842,
+      "learning_rate": 2.9565009560229445e-05,
+      "loss": 1.083,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.12411471224898,
+      "learning_rate": 2.958891013384321e-05,
+      "loss": 1.5391,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.244776713651129,
+      "learning_rate": 2.9612810707456977e-05,
+      "loss": 0.9478,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.34586607427804,
+      "learning_rate": 2.9636711281070743e-05,
+      "loss": 1.4609,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.677980552169689,
+      "learning_rate": 2.9660611854684513e-05,
+      "loss": 0.7007,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.790679608834425,
+      "learning_rate": 2.968451242829828e-05,
+      "loss": 1.2158,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.056673067245094,
+      "learning_rate": 2.970841300191205e-05,
+      "loss": 1.0879,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.042364262744415,
+      "learning_rate": 2.9732313575525815e-05,
+      "loss": 1.7344,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.49986212322597,
+      "learning_rate": 2.9756214149139584e-05,
+      "loss": 1.2441,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.116859213514231,
+      "learning_rate": 2.978011472275335e-05,
+      "loss": 0.8312,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3316142925118233,
+      "learning_rate": 2.9804015296367117e-05,
+      "loss": 0.8872,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.689977169347312,
+      "learning_rate": 2.9827915869980883e-05,
+      "loss": 2.125,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.561561254568156,
+      "learning_rate": 2.985181644359465e-05,
+      "loss": 1.1992,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.186145392563505,
+      "learning_rate": 2.9875717017208415e-05,
+      "loss": 0.7117,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5447860145280905,
+      "learning_rate": 2.989961759082218e-05,
+      "loss": 1.5957,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.827161575537199,
+      "learning_rate": 2.9923518164435948e-05,
+      "loss": 1.291,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.958996836646686,
+      "learning_rate": 2.9947418738049714e-05,
+      "loss": 1.0972,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.580865608810475,
+      "learning_rate": 2.997131931166348e-05,
+      "loss": 1.5469,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.331473515152946,
+      "learning_rate": 2.999521988527725e-05,
+      "loss": 1.293,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.840013293872429,
+      "learning_rate": 3.0019120458891016e-05,
+      "loss": 1.4512,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6261204270544387,
+      "learning_rate": 3.0043021032504782e-05,
+      "loss": 0.9612,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.945749755500083,
+      "learning_rate": 3.0066921606118548e-05,
+      "loss": 1.3291,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.552350724611019,
+      "learning_rate": 3.0090822179732314e-05,
+      "loss": 1.2041,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.982667127074247,
+      "learning_rate": 3.011472275334608e-05,
+      "loss": 1.2529,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.740139623133529,
+      "learning_rate": 3.0138623326959847e-05,
+      "loss": 1.4316,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0711879889469724,
+      "learning_rate": 3.0162523900573613e-05,
+      "loss": 0.9609,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.262089973929399,
+      "learning_rate": 3.018642447418738e-05,
+      "loss": 1.0664,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4025880220058315,
+      "learning_rate": 3.021032504780115e-05,
+      "loss": 1.1992,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6291206879591975,
+      "learning_rate": 3.0234225621414915e-05,
+      "loss": 0.9297,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.987637479164629,
+      "learning_rate": 3.025812619502868e-05,
+      "loss": 1.2402,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.090055939308156,
+      "learning_rate": 3.0282026768642447e-05,
+      "loss": 1.2178,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9809117219248025,
+      "learning_rate": 3.0305927342256213e-05,
+      "loss": 1.2422,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.823565109328541,
+      "learning_rate": 3.0329827915869986e-05,
+      "loss": 1.2715,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.322446989106149,
+      "learning_rate": 3.0353728489483752e-05,
+      "loss": 0.7949,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.925779120303366,
+      "learning_rate": 3.037762906309752e-05,
+      "loss": 1.0615,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4401013328353836,
+      "learning_rate": 3.0401529636711285e-05,
+      "loss": 0.9873,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.255182360317319,
+      "learning_rate": 3.042543021032505e-05,
+      "loss": 1.2656,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.957234792536064,
+      "learning_rate": 3.0449330783938817e-05,
+      "loss": 1.2422,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.934053787686668,
+      "learning_rate": 3.0473231357552583e-05,
+      "loss": 1.3574,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.4305762111420295,
+      "learning_rate": 3.049713193116635e-05,
+      "loss": 0.9863,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7295301840769874,
+      "learning_rate": 3.052103250478012e-05,
+      "loss": 0.8003,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0662601758241332,
+      "learning_rate": 3.0544933078393885e-05,
+      "loss": 1.3574,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6296734117028433,
+      "learning_rate": 3.056883365200765e-05,
+      "loss": 1.4219,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.115846885807436,
+      "learning_rate": 3.059273422562142e-05,
+      "loss": 1.6133,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.95948731850974,
+      "learning_rate": 3.0616634799235184e-05,
+      "loss": 0.9595,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.157793260463166,
+      "learning_rate": 3.064053537284895e-05,
+      "loss": 0.623,
+      "step": 1282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.175287698692353,
+      "learning_rate": 3.0664435946462716e-05,
+      "loss": 0.4009,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.46479417815434,
+      "learning_rate": 3.068833652007648e-05,
+      "loss": 1.3672,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7066706812674464,
+      "learning_rate": 3.071223709369025e-05,
+      "loss": 1.1775,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.598479865627875,
+      "learning_rate": 3.0736137667304014e-05,
+      "loss": 1.1387,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.821738355261675,
+      "learning_rate": 3.076003824091778e-05,
+      "loss": 1.2695,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5764791147816917,
+      "learning_rate": 3.078393881453155e-05,
+      "loss": 1.5469,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.289922909344087,
+      "learning_rate": 3.080783938814531e-05,
+      "loss": 0.9121,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.1982177435863965,
+      "learning_rate": 3.083173996175908e-05,
+      "loss": 1.1387,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.681893564131143,
+      "learning_rate": 3.0855640535372845e-05,
+      "loss": 1.1963,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2070008942533548,
+      "learning_rate": 3.087954110898662e-05,
+      "loss": 1.0381,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.1568534729124185,
+      "learning_rate": 3.0903441682600384e-05,
+      "loss": 1.4727,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.915100872787946,
+      "learning_rate": 3.092734225621415e-05,
+      "loss": 1.6025,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.760632758150895,
+      "learning_rate": 3.095124282982792e-05,
+      "loss": 1.4199,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.160475216153252,
+      "learning_rate": 3.097514340344169e-05,
+      "loss": 1.3809,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.574427642864532,
+      "learning_rate": 3.0999043977055456e-05,
+      "loss": 1.5137,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3861590789051155,
+      "learning_rate": 3.102294455066922e-05,
+      "loss": 1.4102,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0394729297886447,
+      "learning_rate": 3.104684512428299e-05,
+      "loss": 0.8926,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.590581222182695,
+      "learning_rate": 3.1070745697896754e-05,
+      "loss": 0.8623,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7163848235388155,
+      "learning_rate": 3.109464627151052e-05,
+      "loss": 1.0586,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.416321815372709,
+      "learning_rate": 3.111854684512429e-05,
+      "loss": 0.7607,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.832629097740022,
+      "learning_rate": 3.114244741873805e-05,
+      "loss": 1.4727,
+      "step": 1303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.514889911797957,
+      "learning_rate": 3.116634799235182e-05,
+      "loss": 0.9072,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.451646859557162,
+      "learning_rate": 3.1190248565965585e-05,
+      "loss": 1.043,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5344122014271235,
+      "learning_rate": 3.121414913957935e-05,
+      "loss": 0.8672,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8131651197859715,
+      "learning_rate": 3.123804971319312e-05,
+      "loss": 1.5059,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8933519971304382,
+      "learning_rate": 3.1261950286806884e-05,
+      "loss": 1.498,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.762440129879742,
+      "learning_rate": 3.128585086042065e-05,
+      "loss": 0.9805,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.176551668752214,
+      "learning_rate": 3.1309751434034416e-05,
+      "loss": 1.0527,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.829846714027875,
+      "learning_rate": 3.133365200764818e-05,
+      "loss": 0.8291,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.073011210223158,
+      "learning_rate": 3.135755258126195e-05,
+      "loss": 1.2031,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8274280491140331,
+      "learning_rate": 3.1381453154875715e-05,
+      "loss": 1.4141,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.820394439030274,
+      "learning_rate": 3.140535372848948e-05,
+      "loss": 1.6074,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2389910838949283,
+      "learning_rate": 3.1429254302103254e-05,
+      "loss": 0.8311,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.746892012665469,
+      "learning_rate": 3.145315487571702e-05,
+      "loss": 1.1035,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.276111835599268,
+      "learning_rate": 3.1477055449330786e-05,
+      "loss": 1.498,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.201968319372193,
+      "learning_rate": 3.150095602294455e-05,
+      "loss": 0.7661,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.697258111610417,
+      "learning_rate": 3.152485659655832e-05,
+      "loss": 0.7861,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.736831552352838,
+      "learning_rate": 3.1548757170172085e-05,
+      "loss": 1.2168,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.988376830323962,
+      "learning_rate": 3.157265774378586e-05,
+      "loss": 1.2915,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.899814494546712,
+      "learning_rate": 3.1596558317399624e-05,
+      "loss": 1.0713,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.102339732460914,
+      "learning_rate": 3.162045889101339e-05,
+      "loss": 0.9751,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4918022882044264,
+      "learning_rate": 3.1644359464627156e-05,
+      "loss": 1.2695,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6233839181587677,
+      "learning_rate": 3.166826003824092e-05,
+      "loss": 0.9346,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.459734716610438,
+      "learning_rate": 3.169216061185469e-05,
+      "loss": 0.4004,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.596558465415621,
+      "learning_rate": 3.1716061185468455e-05,
+      "loss": 1.2695,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.426245372977716,
+      "learning_rate": 3.173996175908222e-05,
+      "loss": 0.7578,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.451496952224577,
+      "learning_rate": 3.176386233269599e-05,
+      "loss": 1.0977,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.781709536739973,
+      "learning_rate": 3.178776290630975e-05,
+      "loss": 0.8523,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1226239135686544,
+      "learning_rate": 3.181166347992352e-05,
+      "loss": 0.9243,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.372222236548207,
+      "learning_rate": 3.1835564053537285e-05,
+      "loss": 0.9814,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4791148899512456,
+      "learning_rate": 3.185946462715105e-05,
+      "loss": 0.9414,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.330424093416365,
+      "learning_rate": 3.188336520076482e-05,
+      "loss": 1.666,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.9647126448836705,
+      "learning_rate": 3.1907265774378584e-05,
+      "loss": 0.9023,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.042059996003211,
+      "learning_rate": 3.193116634799235e-05,
+      "loss": 0.7891,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9893137231534364,
+      "learning_rate": 3.1955066921606116e-05,
+      "loss": 0.519,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.14984026681899,
+      "learning_rate": 3.197896749521989e-05,
+      "loss": 0.4692,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.356651444108597,
+      "learning_rate": 3.2002868068833655e-05,
+      "loss": 1.2188,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.24484595095206,
+      "learning_rate": 3.202676864244742e-05,
+      "loss": 1.4414,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.902350592774281,
+      "learning_rate": 3.205066921606119e-05,
+      "loss": 1.9453,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1208506693611056,
+      "learning_rate": 3.2074569789674954e-05,
+      "loss": 0.9971,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3907875554287275,
+      "learning_rate": 3.209847036328872e-05,
+      "loss": 1.501,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.883324815508663,
+      "learning_rate": 3.2122370936902486e-05,
+      "loss": 2.3691,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.008180749322298,
+      "learning_rate": 3.214627151051625e-05,
+      "loss": 1.6465,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.134929141714654,
+      "learning_rate": 3.217017208413002e-05,
+      "loss": 0.9912,
+      "step": 1346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.03436230126354,
+      "learning_rate": 3.219407265774379e-05,
+      "loss": 1.4199,
+      "step": 1347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4772315821281063,
+      "learning_rate": 3.221797323135756e-05,
+      "loss": 1.8242,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.387332143530974,
+      "learning_rate": 3.2241873804971324e-05,
+      "loss": 1.582,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.957397209558145,
+      "learning_rate": 3.226577437858509e-05,
+      "loss": 1.4277,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2284933229058046,
+      "learning_rate": 3.2289674952198856e-05,
+      "loss": 1.0669,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.647473234128045,
+      "learning_rate": 3.231357552581262e-05,
+      "loss": 0.6711,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7936595142568645,
+      "learning_rate": 3.233747609942639e-05,
+      "loss": 1.2363,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.778151761186503,
+      "learning_rate": 3.2361376673040155e-05,
+      "loss": 1.6797,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.444653829304766,
+      "learning_rate": 3.238527724665392e-05,
+      "loss": 1.0117,
+      "step": 1355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.835259292189172,
+      "learning_rate": 3.240917782026769e-05,
+      "loss": 0.9082,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5793696075446793,
+      "learning_rate": 3.243307839388145e-05,
+      "loss": 1.25,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.606182585669578,
+      "learning_rate": 3.245697896749522e-05,
+      "loss": 1.0215,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.291843742696321,
+      "learning_rate": 3.2480879541108986e-05,
+      "loss": 1.2461,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.234991125643124,
+      "learning_rate": 3.250478011472275e-05,
+      "loss": 1.9023,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.573091519545894,
+      "learning_rate": 3.2528680688336525e-05,
+      "loss": 1.5586,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.828910627016487,
+      "learning_rate": 3.255258126195029e-05,
+      "loss": 2.1133,
+      "step": 1362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.057566746918535,
+      "learning_rate": 3.257648183556406e-05,
+      "loss": 1.332,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.441203117002124,
+      "learning_rate": 3.260038240917782e-05,
+      "loss": 1.4629,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.906119090777112,
+      "learning_rate": 3.262428298279159e-05,
+      "loss": 1.1973,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.672152410015165,
+      "learning_rate": 3.2648183556405356e-05,
+      "loss": 1.4395,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.571392271674831,
+      "learning_rate": 3.267208413001912e-05,
+      "loss": 1.3008,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.98132449970347,
+      "learning_rate": 3.269598470363289e-05,
+      "loss": 1.8037,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.93627417127746,
+      "learning_rate": 3.2719885277246654e-05,
+      "loss": 1.4629,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.068703421493663,
+      "learning_rate": 3.274378585086042e-05,
+      "loss": 0.7803,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.087868462150265,
+      "learning_rate": 3.2767686424474186e-05,
+      "loss": 0.9736,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3420602682399148,
+      "learning_rate": 3.279158699808795e-05,
+      "loss": 1.1504,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.174836228119242,
+      "learning_rate": 3.2815487571701726e-05,
+      "loss": 1.3359,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7105127710950354,
+      "learning_rate": 3.283938814531549e-05,
+      "loss": 1.4082,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7711932491123354,
+      "learning_rate": 3.286328871892926e-05,
+      "loss": 1.3652,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4632837161830046,
+      "learning_rate": 3.2887189292543024e-05,
+      "loss": 0.6108,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.35357681131564,
+      "learning_rate": 3.291108986615679e-05,
+      "loss": 0.8655,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.985954306179401,
+      "learning_rate": 3.2934990439770556e-05,
+      "loss": 1.0039,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.467634004342692,
+      "learning_rate": 3.295889101338432e-05,
+      "loss": 0.7761,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.705312976694051,
+      "learning_rate": 3.298279158699809e-05,
+      "loss": 1.4053,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.957763786927131,
+      "learning_rate": 3.3006692160611855e-05,
+      "loss": 1.2754,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.890835481146019,
+      "learning_rate": 3.303059273422562e-05,
+      "loss": 0.8335,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6632940792055155,
+      "learning_rate": 3.305449330783939e-05,
+      "loss": 1.2959,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.39817755500663,
+      "learning_rate": 3.307839388145316e-05,
+      "loss": 0.8574,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.657326972151068,
+      "learning_rate": 3.3102294455066926e-05,
+      "loss": 1.5879,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9200025114967945,
+      "learning_rate": 3.312619502868069e-05,
+      "loss": 0.9326,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.876762302983034,
+      "learning_rate": 3.315009560229446e-05,
+      "loss": 0.9121,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8317209218444224,
+      "learning_rate": 3.3173996175908225e-05,
+      "loss": 1.0801,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8588425736217338,
+      "learning_rate": 3.319789674952199e-05,
+      "loss": 1.0527,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.403255233906569,
+      "learning_rate": 3.322179732313576e-05,
+      "loss": 1.1777,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.534510265970042,
+      "learning_rate": 3.324569789674952e-05,
+      "loss": 1.8906,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.029426576317721,
+      "learning_rate": 3.326959847036329e-05,
+      "loss": 1.4226,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.013373419182051,
+      "learning_rate": 3.3293499043977056e-05,
+      "loss": 0.9033,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6634353391698324,
+      "learning_rate": 3.331739961759082e-05,
+      "loss": 1.0835,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2793188315878257,
+      "learning_rate": 3.334130019120459e-05,
+      "loss": 1.0732,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.279836602345169,
+      "learning_rate": 3.3365200764818354e-05,
+      "loss": 1.1924,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.436051940999221,
+      "learning_rate": 3.338910133843212e-05,
+      "loss": 1.2109,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.481939213348817,
+      "learning_rate": 3.3413001912045887e-05,
+      "loss": 1.2832,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.33654279376666,
+      "learning_rate": 3.343690248565965e-05,
+      "loss": 0.4399,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2840783418053148,
+      "learning_rate": 3.3460803059273426e-05,
+      "loss": 0.8418,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8496492002657443,
+      "learning_rate": 3.348470363288719e-05,
+      "loss": 0.5662,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.583072423577546,
+      "learning_rate": 3.350860420650096e-05,
+      "loss": 1.1777,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.069941918137427,
+      "learning_rate": 3.3532504780114724e-05,
+      "loss": 1.0488,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.900432383432556,
+      "learning_rate": 3.355640535372849e-05,
+      "loss": 1.1543,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5139355406346082,
+      "learning_rate": 3.3580305927342257e-05,
+      "loss": 0.8037,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.444036819703003,
+      "learning_rate": 3.360420650095602e-05,
+      "loss": 0.8242,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.700674609482837,
+      "learning_rate": 3.3628107074569796e-05,
+      "loss": 1.5566,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.608010093201074,
+      "learning_rate": 3.365200764818356e-05,
+      "loss": 1.377,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.586470451874787,
+      "learning_rate": 3.367590822179733e-05,
+      "loss": 1.2578,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.125260509289839,
+      "learning_rate": 3.3699808795411094e-05,
+      "loss": 0.7178,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3199271662892444,
+      "learning_rate": 3.372370936902486e-05,
+      "loss": 0.9346,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5749000108514153,
+      "learning_rate": 3.3747609942638627e-05,
+      "loss": 1.5078,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.173225830931043,
+      "learning_rate": 3.377151051625239e-05,
+      "loss": 1.8027,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.783159014147273,
+      "learning_rate": 3.379541108986616e-05,
+      "loss": 0.7656,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.868584152354725,
+      "learning_rate": 3.3819311663479925e-05,
+      "loss": 0.9814,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.44840065746737,
+      "learning_rate": 3.384321223709369e-05,
+      "loss": 0.9658,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5573535353426764,
+      "learning_rate": 3.386711281070746e-05,
+      "loss": 0.4673,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 31.334750312527426,
+      "learning_rate": 3.3891013384321224e-05,
+      "loss": 1.0029,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.050154338026772,
+      "learning_rate": 3.391491395793499e-05,
+      "loss": 1.1172,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6920553235835207,
+      "learning_rate": 3.3938814531548756e-05,
+      "loss": 1.2012,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.957877397102365,
+      "learning_rate": 3.396271510516252e-05,
+      "loss": 1.7383,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.006739358072597,
+      "learning_rate": 3.398661567877629e-05,
+      "loss": 0.9287,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.740718524167894,
+      "learning_rate": 3.4010516252390054e-05,
+      "loss": 1.2012,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.418564922336254,
+      "learning_rate": 3.403441682600382e-05,
+      "loss": 0.521,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.029752987338053,
+      "learning_rate": 3.405831739961759e-05,
+      "loss": 1.2373,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.722244273466298,
+      "learning_rate": 3.408221797323136e-05,
+      "loss": 1.4043,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2678386528711743,
+      "learning_rate": 3.4106118546845126e-05,
+      "loss": 1.3633,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.163101288970541,
+      "learning_rate": 3.413001912045889e-05,
+      "loss": 1.3516,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6936873249454054,
+      "learning_rate": 3.415391969407266e-05,
+      "loss": 1.6055,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.162262436124673,
+      "learning_rate": 3.417782026768643e-05,
+      "loss": 1.5918,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.339349007352738,
+      "learning_rate": 3.42017208413002e-05,
+      "loss": 1.0332,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.136077835376979,
+      "learning_rate": 3.4225621414913964e-05,
+      "loss": 0.7383,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.908416432793868,
+      "learning_rate": 3.424952198852773e-05,
+      "loss": 1.2578,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.898781916600818,
+      "learning_rate": 3.4273422562141496e-05,
+      "loss": 1.6973,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.28367400770443,
+      "learning_rate": 3.429732313575526e-05,
+      "loss": 0.9941,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4329129609305524,
+      "learning_rate": 3.432122370936903e-05,
+      "loss": 0.751,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.066601069343631,
+      "learning_rate": 3.4345124282982794e-05,
+      "loss": 0.9912,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4589523786018983,
+      "learning_rate": 3.436902485659656e-05,
+      "loss": 1.5723,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.911983778085462,
+      "learning_rate": 3.439292543021033e-05,
+      "loss": 0.8633,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.719521424040978,
+      "learning_rate": 3.441682600382409e-05,
+      "loss": 1.1172,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.734832238001837,
+      "learning_rate": 3.444072657743786e-05,
+      "loss": 0.9648,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.783146158271168,
+      "learning_rate": 3.4464627151051625e-05,
+      "loss": 1.3809,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.573976270860137,
+      "learning_rate": 3.448852772466539e-05,
+      "loss": 1.0488,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.4121499034254805,
+      "learning_rate": 3.451242829827916e-05,
+      "loss": 1.2402,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.993944697589083,
+      "learning_rate": 3.4536328871892924e-05,
+      "loss": 1.0449,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9147292287184423,
+      "learning_rate": 3.456022944550669e-05,
+      "loss": 0.8513,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.049991541724144,
+      "learning_rate": 3.4584130019120456e-05,
+      "loss": 1.2939,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.194575436731621,
+      "learning_rate": 3.460803059273422e-05,
+      "loss": 1.4688,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.459575491917951,
+      "learning_rate": 3.463193116634799e-05,
+      "loss": 0.585,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.956622215812216,
+      "learning_rate": 3.465583173996176e-05,
+      "loss": 0.9297,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.501209520339627,
+      "learning_rate": 3.467973231357553e-05,
+      "loss": 0.8467,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.42011929274749,
+      "learning_rate": 3.4703632887189294e-05,
+      "loss": 1.4219,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.999167801571183,
+      "learning_rate": 3.472753346080307e-05,
+      "loss": 1.3906,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.2014849822528335,
+      "learning_rate": 3.475143403441683e-05,
+      "loss": 1.0938,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.843437261467167,
+      "learning_rate": 3.47753346080306e-05,
+      "loss": 1.3438,
+      "step": 1455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.08515102228911,
+      "learning_rate": 3.4799235181644365e-05,
+      "loss": 1.6436,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.369208132491211,
+      "learning_rate": 3.482313575525813e-05,
+      "loss": 1.6465,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.126114280267135,
+      "learning_rate": 3.48470363288719e-05,
+      "loss": 1.4219,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7366222629204984,
+      "learning_rate": 3.4870936902485664e-05,
+      "loss": 1.5059,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7399689048896088,
+      "learning_rate": 3.489483747609943e-05,
+      "loss": 1.3789,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.360388008130029,
+      "learning_rate": 3.4918738049713196e-05,
+      "loss": 1.5547,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.283594533672188,
+      "learning_rate": 3.494263862332696e-05,
+      "loss": 1.1641,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.58990742596681,
+      "learning_rate": 3.496653919694073e-05,
+      "loss": 1.3145,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0916680135688726,
+      "learning_rate": 3.4990439770554495e-05,
+      "loss": 0.8571,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.158425623344375,
+      "learning_rate": 3.501434034416826e-05,
+      "loss": 1.1523,
+      "step": 1465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.554339348726577,
+      "learning_rate": 3.503824091778203e-05,
+      "loss": 1.6748,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.340419967157532,
+      "learning_rate": 3.506214149139579e-05,
+      "loss": 1.9922,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.352896541750713,
+      "learning_rate": 3.508604206500956e-05,
+      "loss": 1.5469,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.079022117047168,
+      "learning_rate": 3.5109942638623325e-05,
+      "loss": 1.0264,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9106821698649488,
+      "learning_rate": 3.513384321223709e-05,
+      "loss": 1.0039,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.606596338716999,
+      "learning_rate": 3.515774378585086e-05,
+      "loss": 0.7373,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.819942503656011,
+      "learning_rate": 3.5181644359464624e-05,
+      "loss": 1.1533,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.62606819807778,
+      "learning_rate": 3.52055449330784e-05,
+      "loss": 1.7539,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.592836408394804,
+      "learning_rate": 3.522944550669216e-05,
+      "loss": 1.1699,
+      "step": 1474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7851032610503212,
+      "learning_rate": 3.525334608030593e-05,
+      "loss": 0.9766,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.797824504237222,
+      "learning_rate": 3.5277246653919695e-05,
+      "loss": 1.25,
+      "step": 1476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.593945572548697,
+      "learning_rate": 3.530114722753346e-05,
+      "loss": 1.3037,
+      "step": 1477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6006242104946464,
+      "learning_rate": 3.5325047801147234e-05,
+      "loss": 1.0811,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.272076488806982,
+      "learning_rate": 3.5348948374761e-05,
+      "loss": 1.1508,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.21153288229702,
+      "learning_rate": 3.537284894837477e-05,
+      "loss": 1.1523,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.056871445557164,
+      "learning_rate": 3.539674952198853e-05,
+      "loss": 1.0156,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.892612424585462,
+      "learning_rate": 3.54206500956023e-05,
+      "loss": 0.8135,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.368935540633112,
+      "learning_rate": 3.5444550669216065e-05,
+      "loss": 1.2578,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.257325647268369,
+      "learning_rate": 3.546845124282983e-05,
+      "loss": 1.1035,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9644247104958712,
+      "learning_rate": 3.54923518164436e-05,
+      "loss": 0.4746,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7605913118980334,
+      "learning_rate": 3.5516252390057364e-05,
+      "loss": 0.728,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.979786385586607,
+      "learning_rate": 3.554015296367113e-05,
+      "loss": 1.998,
+      "step": 1487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7446652599859567,
+      "learning_rate": 3.5564053537284896e-05,
+      "loss": 1.5254,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.276062209915951,
+      "learning_rate": 3.558795411089866e-05,
+      "loss": 1.7012,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8140118212304155,
+      "learning_rate": 3.561185468451243e-05,
+      "loss": 0.7275,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.763608011628596,
+      "learning_rate": 3.5635755258126195e-05,
+      "loss": 0.7764,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.232628220445203,
+      "learning_rate": 3.565965583173996e-05,
+      "loss": 1.5234,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.646668567681772,
+      "learning_rate": 3.568355640535373e-05,
+      "loss": 1.5479,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.626039844004387,
+      "learning_rate": 3.570745697896749e-05,
+      "loss": 1.2617,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2884557168817237,
+      "learning_rate": 3.573135755258126e-05,
+      "loss": 1.2051,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.969628394237216,
+      "learning_rate": 3.575525812619503e-05,
+      "loss": 1.1426,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2023725416681117,
+      "learning_rate": 3.57791586998088e-05,
+      "loss": 1.1602,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.30613560637515,
+      "learning_rate": 3.5803059273422565e-05,
+      "loss": 1.3398,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.174440106788973,
+      "learning_rate": 3.582695984703633e-05,
+      "loss": 1.1504,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.235209923938959,
+      "learning_rate": 3.58508604206501e-05,
+      "loss": 1.2539,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1533550787301134,
+      "learning_rate": 3.587476099426386e-05,
+      "loss": 1.4561,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.231054827247205,
+      "learning_rate": 3.589866156787763e-05,
+      "loss": 0.9141,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.869513689266457,
+      "learning_rate": 3.5922562141491396e-05,
+      "loss": 1.144,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9011893562970528,
+      "learning_rate": 3.594646271510517e-05,
+      "loss": 0.8027,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.322540758030975,
+      "learning_rate": 3.5970363288718935e-05,
+      "loss": 1.8906,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.485438204210271,
+      "learning_rate": 3.59942638623327e-05,
+      "loss": 0.7095,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1404748377794838,
+      "learning_rate": 3.601816443594647e-05,
+      "loss": 1.207,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.969826432881705,
+      "learning_rate": 3.604206500956023e-05,
+      "loss": 0.7783,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.994242388345268,
+      "learning_rate": 3.6065965583174e-05,
+      "loss": 1.3691,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.640109412429377,
+      "learning_rate": 3.6089866156787766e-05,
+      "loss": 1.3955,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4496518443148525,
+      "learning_rate": 3.611376673040153e-05,
+      "loss": 1.6797,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.019674162055057,
+      "learning_rate": 3.61376673040153e-05,
+      "loss": 1.5605,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2511958515472905,
+      "learning_rate": 3.6161567877629064e-05,
+      "loss": 0.8291,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2349084976203075,
+      "learning_rate": 3.618546845124283e-05,
+      "loss": 0.6924,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.701330710795669,
+      "learning_rate": 3.6209369024856596e-05,
+      "loss": 1.2676,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.833518265317465,
+      "learning_rate": 3.623326959847036e-05,
+      "loss": 1.3076,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0400269848479167,
+      "learning_rate": 3.625717017208413e-05,
+      "loss": 1.3945,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.487325488850626,
+      "learning_rate": 3.6281070745697895e-05,
+      "loss": 1.2871,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.430272408542346,
+      "learning_rate": 3.630497131931167e-05,
+      "loss": 0.9897,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3831548715763518,
+      "learning_rate": 3.6328871892925434e-05,
+      "loss": 1.1934,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.615329801514826,
+      "learning_rate": 3.63527724665392e-05,
+      "loss": 1.5625,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2574880822290626,
+      "learning_rate": 3.6376673040152966e-05,
+      "loss": 0.5483,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.611693959506361,
+      "learning_rate": 3.640057361376673e-05,
+      "loss": 1.2012,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5650273397078025,
+      "learning_rate": 3.64244741873805e-05,
+      "loss": 1.4062,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.328251686415722,
+      "learning_rate": 3.6448374760994265e-05,
+      "loss": 1.3486,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.129813376261222,
+      "learning_rate": 3.647227533460803e-05,
+      "loss": 0.4284,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.045847881920957,
+      "learning_rate": 3.64961759082218e-05,
+      "loss": 0.8096,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.567095560502022,
+      "learning_rate": 3.652007648183556e-05,
+      "loss": 1.3516,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4051756008383087,
+      "learning_rate": 3.654397705544933e-05,
+      "loss": 1.1177,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3922850386495003,
+      "learning_rate": 3.65678776290631e-05,
+      "loss": 1.4014,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.109486161385644,
+      "learning_rate": 3.659177820267687e-05,
+      "loss": 0.9951,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.901024960671603,
+      "learning_rate": 3.6615678776290635e-05,
+      "loss": 1.1699,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.292025266391569,
+      "learning_rate": 3.66395793499044e-05,
+      "loss": 0.9961,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.845359775294457,
+      "learning_rate": 3.666347992351817e-05,
+      "loss": 1.0908,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9102232571128965,
+      "learning_rate": 3.668738049713193e-05,
+      "loss": 0.9419,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7380460308377836,
+      "learning_rate": 3.67112810707457e-05,
+      "loss": 1.3965,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.061321850649188,
+      "learning_rate": 3.6735181644359466e-05,
+      "loss": 0.5864,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4665610350939526,
+      "learning_rate": 3.675908221797323e-05,
+      "loss": 1.5137,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9801195427421028,
+      "learning_rate": 3.6782982791587e-05,
+      "loss": 1.3086,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.352971011667544,
+      "learning_rate": 3.6806883365200764e-05,
+      "loss": 1.0879,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.02707392320704,
+      "learning_rate": 3.683078393881453e-05,
+      "loss": 0.916,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.2920318228079966,
+      "learning_rate": 3.68546845124283e-05,
+      "loss": 1.8096,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3806585864048615,
+      "learning_rate": 3.687858508604207e-05,
+      "loss": 1.0889,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.502897561795117,
+      "learning_rate": 3.6902485659655836e-05,
+      "loss": 0.9805,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.634548948988595,
+      "learning_rate": 3.69263862332696e-05,
+      "loss": 1.0029,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.375942946209549,
+      "learning_rate": 3.695028680688337e-05,
+      "loss": 0.8956,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8797151734328814,
+      "learning_rate": 3.6974187380497134e-05,
+      "loss": 0.8359,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3955674954759782,
+      "learning_rate": 3.69980879541109e-05,
+      "loss": 1.0869,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.08969634775143,
+      "learning_rate": 3.7021988527724667e-05,
+      "loss": 0.2537,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.278616233996742,
+      "learning_rate": 3.704588910133843e-05,
+      "loss": 0.9668,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6548358518788207,
+      "learning_rate": 3.70697896749522e-05,
+      "loss": 1.3325,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.833014817540135,
+      "learning_rate": 3.7093690248565965e-05,
+      "loss": 0.791,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.604753433244315,
+      "learning_rate": 3.711759082217973e-05,
+      "loss": 0.5278,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.875588729951354,
+      "learning_rate": 3.71414913957935e-05,
+      "loss": 0.7109,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.949014995225207,
+      "learning_rate": 3.7165391969407264e-05,
+      "loss": 0.8545,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4369219559173776,
+      "learning_rate": 3.7189292543021036e-05,
+      "loss": 0.8755,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.951790159784381,
+      "learning_rate": 3.72131931166348e-05,
+      "loss": 1.3008,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.691244810472433,
+      "learning_rate": 3.723709369024857e-05,
+      "loss": 1.2188,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.911226936574824,
+      "learning_rate": 3.7260994263862335e-05,
+      "loss": 1.2422,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.655738956133724,
+      "learning_rate": 3.72848948374761e-05,
+      "loss": 1.4824,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5260842440972486,
+      "learning_rate": 3.730879541108987e-05,
+      "loss": 0.9385,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.303021756450235,
+      "learning_rate": 3.7332695984703634e-05,
+      "loss": 1.7168,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0758075766859,
+      "learning_rate": 3.73565965583174e-05,
+      "loss": 1.2461,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.008788205699137,
+      "learning_rate": 3.7380497131931166e-05,
+      "loss": 0.8252,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.953161524863346,
+      "learning_rate": 3.740439770554494e-05,
+      "loss": 1.2207,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.860931909658119,
+      "learning_rate": 3.7428298279158705e-05,
+      "loss": 1.9219,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.4461467875501635,
+      "learning_rate": 3.745219885277247e-05,
+      "loss": 1.2637,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.176342650862261,
+      "learning_rate": 3.747609942638624e-05,
+      "loss": 1.7129,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.068360312649916,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.8555,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1562611796870383,
+      "learning_rate": 3.752390057361377e-05,
+      "loss": 0.8438,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.496733093760305,
+      "learning_rate": 3.7547801147227536e-05,
+      "loss": 1.208,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5305540281873444,
+      "learning_rate": 3.75717017208413e-05,
+      "loss": 1.5586,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5887759076402066,
+      "learning_rate": 3.759560229445507e-05,
+      "loss": 1.1279,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2502856495843626,
+      "learning_rate": 3.7619502868068834e-05,
+      "loss": 0.8274,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4547997850248615,
+      "learning_rate": 3.76434034416826e-05,
+      "loss": 1.3535,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.073673083740088,
+      "learning_rate": 3.766730401529637e-05,
+      "loss": 1.1777,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.990460068587564,
+      "learning_rate": 3.769120458891013e-05,
+      "loss": 1.2148,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9651891936848993,
+      "learning_rate": 3.77151051625239e-05,
+      "loss": 1.1826,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.153105134911561,
+      "learning_rate": 3.7739005736137665e-05,
+      "loss": 1.5898,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.440130612391117,
+      "learning_rate": 3.776290630975143e-05,
+      "loss": 1.5,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.69426417249553,
+      "learning_rate": 3.77868068833652e-05,
+      "loss": 1.1064,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.628877772818924,
+      "learning_rate": 3.7810707456978964e-05,
+      "loss": 1.627,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6857850564348933,
+      "learning_rate": 3.783460803059274e-05,
+      "loss": 1.375,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3003246327056797,
+      "learning_rate": 3.78585086042065e-05,
+      "loss": 1.166,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.756216364502067,
+      "learning_rate": 3.788240917782027e-05,
+      "loss": 1.1689,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.243214441845561,
+      "learning_rate": 3.7906309751434035e-05,
+      "loss": 1.6602,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.364530843754456,
+      "learning_rate": 3.79302103250478e-05,
+      "loss": 1.3535,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.449781802539307,
+      "learning_rate": 3.7954110898661574e-05,
+      "loss": 1.1777,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.903486570643467,
+      "learning_rate": 3.797801147227534e-05,
+      "loss": 1.1758,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.769896051810325,
+      "learning_rate": 3.800191204588911e-05,
+      "loss": 1.0059,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5470503036937036,
+      "learning_rate": 3.802581261950287e-05,
+      "loss": 1.3262,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.280887481083015,
+      "learning_rate": 3.804971319311664e-05,
+      "loss": 1.4355,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.811020859147601,
+      "learning_rate": 3.8073613766730405e-05,
+      "loss": 0.7432,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.27661665601337,
+      "learning_rate": 3.809751434034417e-05,
+      "loss": 1.5488,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.118569127649837,
+      "learning_rate": 3.812141491395794e-05,
+      "loss": 1.4023,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.698605678952305,
+      "learning_rate": 3.8145315487571704e-05,
+      "loss": 1.0977,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5739896958922204,
+      "learning_rate": 3.816921606118547e-05,
+      "loss": 1.1025,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.529618804926562,
+      "learning_rate": 3.8193116634799236e-05,
+      "loss": 1.458,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.777130546397514,
+      "learning_rate": 3.8217017208413e-05,
+      "loss": 1.0928,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.53623044280822,
+      "learning_rate": 3.824091778202677e-05,
+      "loss": 0.9619,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6157923864921164,
+      "learning_rate": 3.8264818355640535e-05,
+      "loss": 1.2441,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.37466581975651,
+      "learning_rate": 3.82887189292543e-05,
+      "loss": 1.1084,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.085368352628082,
+      "learning_rate": 3.831261950286807e-05,
+      "loss": 0.8857,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.105577638074166,
+      "learning_rate": 3.833652007648183e-05,
+      "loss": 1.332,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0920370126152403,
+      "learning_rate": 3.83604206500956e-05,
+      "loss": 1.2334,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.878474085344101,
+      "learning_rate": 3.8384321223709365e-05,
+      "loss": 1.3203,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0292641338494986,
+      "learning_rate": 3.840822179732314e-05,
+      "loss": 0.7222,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.523261453895612,
+      "learning_rate": 3.8432122370936904e-05,
+      "loss": 1.2383,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.00982235287083,
+      "learning_rate": 3.845602294455067e-05,
+      "loss": 0.9399,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.504454984620848,
+      "learning_rate": 3.847992351816444e-05,
+      "loss": 1.752,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.066141571479135,
+      "learning_rate": 3.850382409177821e-05,
+      "loss": 0.6875,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.500352328026076,
+      "learning_rate": 3.8527724665391976e-05,
+      "loss": 0.9375,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.183237553699998,
+      "learning_rate": 3.855162523900574e-05,
+      "loss": 0.9883,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5488273766761353,
+      "learning_rate": 3.857552581261951e-05,
+      "loss": 1.3901,
+      "step": 1614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.521850307478905,
+      "learning_rate": 3.8599426386233274e-05,
+      "loss": 1.0596,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1635500645999692,
+      "learning_rate": 3.862332695984704e-05,
+      "loss": 1.8691,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.02165415954214,
+      "learning_rate": 3.864722753346081e-05,
+      "loss": 1.7393,
+      "step": 1617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.514043046198681,
+      "learning_rate": 3.867112810707457e-05,
+      "loss": 1.1992,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5131557354806695,
+      "learning_rate": 3.869502868068834e-05,
+      "loss": 1.1904,
+      "step": 1619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.258111191763579,
+      "learning_rate": 3.8718929254302105e-05,
+      "loss": 1.4648,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9020138292462905,
+      "learning_rate": 3.874282982791587e-05,
+      "loss": 1.457,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7081367030046035,
+      "learning_rate": 3.876673040152964e-05,
+      "loss": 0.8455,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0824822118884785,
+      "learning_rate": 3.8790630975143404e-05,
+      "loss": 0.6445,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2150503126160936,
+      "learning_rate": 3.881453154875717e-05,
+      "loss": 1.01,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3999410383611224,
+      "learning_rate": 3.8838432122370936e-05,
+      "loss": 0.7905,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.6393471959075026,
+      "learning_rate": 3.88623326959847e-05,
+      "loss": 1.2305,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.38524605003089,
+      "learning_rate": 3.888623326959847e-05,
+      "loss": 0.7007,
+      "step": 1627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.946901390488832,
+      "learning_rate": 3.8910133843212235e-05,
+      "loss": 1.2021,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6168061391345008,
+      "learning_rate": 3.8934034416826e-05,
+      "loss": 1.3203,
+      "step": 1629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1207056675574822,
+      "learning_rate": 3.8957934990439774e-05,
+      "loss": 1.2734,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0224045875026486,
+      "learning_rate": 3.898183556405354e-05,
+      "loss": 1.4453,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.7150395544772445,
+      "learning_rate": 3.9005736137667306e-05,
+      "loss": 1.6758,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.208640678189342,
+      "learning_rate": 3.902963671128107e-05,
+      "loss": 0.8203,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5266497870355584,
+      "learning_rate": 3.905353728489484e-05,
+      "loss": 0.9985,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9202655162604059,
+      "learning_rate": 3.907743785850861e-05,
+      "loss": 0.7627,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2358895883818426,
+      "learning_rate": 3.910133843212238e-05,
+      "loss": 0.8389,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.038140374292989,
+      "learning_rate": 3.9125239005736144e-05,
+      "loss": 1.0806,
+      "step": 1637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.008662380953609,
+      "learning_rate": 3.914913957934991e-05,
+      "loss": 1.0928,
+      "step": 1638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.205298453969008,
+      "learning_rate": 3.9173040152963676e-05,
+      "loss": 1.1816,
+      "step": 1639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.146080694169311,
+      "learning_rate": 3.919694072657744e-05,
+      "loss": 0.8125,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.160035044408526,
+      "learning_rate": 3.922084130019121e-05,
+      "loss": 1.2363,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.250353301739876,
+      "learning_rate": 3.9244741873804975e-05,
+      "loss": 1.3828,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6483016527832555,
+      "learning_rate": 3.926864244741874e-05,
+      "loss": 1.3555,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.348286626722619,
+      "learning_rate": 3.929254302103251e-05,
+      "loss": 1.1855,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8600712816050775,
+      "learning_rate": 3.931644359464627e-05,
+      "loss": 1.0586,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.552151182945554,
+      "learning_rate": 3.934034416826004e-05,
+      "loss": 1.2715,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.642155534300417,
+      "learning_rate": 3.9364244741873806e-05,
+      "loss": 1.0527,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6506821564113694,
+      "learning_rate": 3.938814531548757e-05,
+      "loss": 1.0996,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.962465643185107,
+      "learning_rate": 3.941204588910134e-05,
+      "loss": 1.9297,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.332335175945567,
+      "learning_rate": 3.9435946462715104e-05,
+      "loss": 1.1387,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6595729014615253,
+      "learning_rate": 3.945984703632887e-05,
+      "loss": 1.208,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.452035228497327,
+      "learning_rate": 3.9483747609942636e-05,
+      "loss": 1.1992,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8005262016673913,
+      "learning_rate": 3.950764818355641e-05,
+      "loss": 0.9209,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.293088566736569,
+      "learning_rate": 3.9531548757170175e-05,
+      "loss": 1.0391,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8771211843097784,
+      "learning_rate": 3.955544933078394e-05,
+      "loss": 1.1274,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.214886763552617,
+      "learning_rate": 3.957934990439771e-05,
+      "loss": 0.6191,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0920729281965738,
+      "learning_rate": 3.9603250478011474e-05,
+      "loss": 1.2842,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7626174694649226,
+      "learning_rate": 3.962715105162524e-05,
+      "loss": 1.8125,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.897249273881643,
+      "learning_rate": 3.9651051625239006e-05,
+      "loss": 1.2852,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9251548153419193,
+      "learning_rate": 3.967495219885277e-05,
+      "loss": 0.7505,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.40608032923895,
+      "learning_rate": 3.9698852772466545e-05,
+      "loss": 0.8877,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.548569049781097,
+      "learning_rate": 3.972275334608031e-05,
+      "loss": 1.9199,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.482820035601987,
+      "learning_rate": 3.974665391969408e-05,
+      "loss": 1.0811,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.557490126813548,
+      "learning_rate": 3.9770554493307844e-05,
+      "loss": 0.5586,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.626926346582411,
+      "learning_rate": 3.979445506692161e-05,
+      "loss": 1.1445,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.97445370535781,
+      "learning_rate": 3.9818355640535376e-05,
+      "loss": 1.5117,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8914836536138693,
+      "learning_rate": 3.984225621414914e-05,
+      "loss": 0.8428,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.244386517190797,
+      "learning_rate": 3.986615678776291e-05,
+      "loss": 1.3809,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.595112501226252,
+      "learning_rate": 3.9890057361376675e-05,
+      "loss": 1.2402,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1628267865724724,
+      "learning_rate": 3.991395793499044e-05,
+      "loss": 1.5918,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.428663513672672,
+      "learning_rate": 3.993785850860421e-05,
+      "loss": 1.0938,
+      "step": 1671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3983910568396216,
+      "learning_rate": 3.996175908221797e-05,
+      "loss": 0.9199,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.175548821846803,
+      "learning_rate": 3.998565965583174e-05,
+      "loss": 0.7473,
+      "step": 1673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.416969992795823,
+      "learning_rate": 4.0009560229445506e-05,
+      "loss": 0.9272,
+      "step": 1674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8112027822380257,
+      "learning_rate": 4.003346080305927e-05,
+      "loss": 1.3672,
+      "step": 1675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4834651015575755,
+      "learning_rate": 4.0057361376673045e-05,
+      "loss": 1.209,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.851392432589114,
+      "learning_rate": 4.008126195028681e-05,
+      "loss": 0.9795,
+      "step": 1677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.482043437789757,
+      "learning_rate": 4.010516252390058e-05,
+      "loss": 1.3242,
+      "step": 1678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.57977869724376,
+      "learning_rate": 4.012906309751434e-05,
+      "loss": 1.3271,
+      "step": 1679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.505895728470933,
+      "learning_rate": 4.015296367112811e-05,
+      "loss": 1.5898,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.867542782969348,
+      "learning_rate": 4.0176864244741876e-05,
+      "loss": 1.5303,
+      "step": 1681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.860242701549569,
+      "learning_rate": 4.020076481835564e-05,
+      "loss": 0.9951,
+      "step": 1682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8551389240195757,
+      "learning_rate": 4.022466539196941e-05,
+      "loss": 1.1328,
+      "step": 1683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.382913680107162,
+      "learning_rate": 4.0248565965583174e-05,
+      "loss": 0.9131,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.4204446186536765,
+      "learning_rate": 4.027246653919694e-05,
+      "loss": 1.6309,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.9657614051407055,
+      "learning_rate": 4.0296367112810707e-05,
+      "loss": 1.8496,
+      "step": 1686
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.301351220817639,
+      "learning_rate": 4.032026768642448e-05,
+      "loss": 1.1523,
+      "step": 1687
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.604637917159388,
+      "learning_rate": 4.0344168260038246e-05,
+      "loss": 1.8145,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.27732212287626,
+      "learning_rate": 4.036806883365201e-05,
+      "loss": 0.8975,
+      "step": 1689
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.816202894188861,
+      "learning_rate": 4.039196940726578e-05,
+      "loss": 1.498,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.15082350683718,
+      "learning_rate": 4.0415869980879544e-05,
+      "loss": 1.1191,
+      "step": 1691
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.469596717591818,
+      "learning_rate": 4.043977055449331e-05,
+      "loss": 0.6206,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.9915116984665335,
+      "learning_rate": 4.0463671128107076e-05,
+      "loss": 1.1191,
+      "step": 1693
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.823064705929987,
+      "learning_rate": 4.048757170172084e-05,
+      "loss": 1.6152,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.419104289443079,
+      "learning_rate": 4.051147227533461e-05,
+      "loss": 0.9192,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.352415745392575,
+      "learning_rate": 4.0535372848948375e-05,
+      "loss": 1.1113,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.745219425959444,
+      "learning_rate": 4.055927342256214e-05,
+      "loss": 1.0146,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.810174945978442,
+      "learning_rate": 4.058317399617591e-05,
+      "loss": 1.6826,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1202311017373012,
+      "learning_rate": 4.060707456978968e-05,
+      "loss": 0.8091,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2090842288937766,
+      "learning_rate": 4.0630975143403446e-05,
+      "loss": 0.998,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5027221164965803,
+      "learning_rate": 4.065487571701721e-05,
+      "loss": 1.1709,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.142492625555553,
+      "learning_rate": 4.067877629063098e-05,
+      "loss": 1.2676,
+      "step": 1702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.846528665256135,
+      "learning_rate": 4.0702676864244745e-05,
+      "loss": 0.7183,
+      "step": 1703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7134587329533266,
+      "learning_rate": 4.072657743785851e-05,
+      "loss": 1.5332,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8915547293362063,
+      "learning_rate": 4.075047801147228e-05,
+      "loss": 1.0146,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.725561707025715,
+      "learning_rate": 4.0774378585086043e-05,
+      "loss": 0.9155,
+      "step": 1706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.341066634844726,
+      "learning_rate": 4.079827915869981e-05,
+      "loss": 1.1719,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.345732438426001,
+      "learning_rate": 4.0822179732313576e-05,
+      "loss": 1.9355,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.596753634046127,
+      "learning_rate": 4.084608030592734e-05,
+      "loss": 2.1602,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.972587592357792,
+      "learning_rate": 4.086998087954111e-05,
+      "loss": 1.4551,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.893897768325498,
+      "learning_rate": 4.0893881453154874e-05,
+      "loss": 0.6643,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4754437829008045,
+      "learning_rate": 4.091778202676864e-05,
+      "loss": 1.1523,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4944314254601783,
+      "learning_rate": 4.0941682600382413e-05,
+      "loss": 1.4951,
+      "step": 1713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.562449765782212,
+      "learning_rate": 4.096558317399618e-05,
+      "loss": 0.8662,
+      "step": 1714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9615341028339577,
+      "learning_rate": 4.0989483747609946e-05,
+      "loss": 0.8896,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.005692368252947,
+      "learning_rate": 4.101338432122371e-05,
+      "loss": 0.8311,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.715338397074335,
+      "learning_rate": 4.103728489483748e-05,
+      "loss": 1.5508,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.09087148321571,
+      "learning_rate": 4.1061185468451244e-05,
+      "loss": 1.4766,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.761446797795174,
+      "learning_rate": 4.108508604206501e-05,
+      "loss": 1.3418,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5258506073269738,
+      "learning_rate": 4.110898661567878e-05,
+      "loss": 0.9131,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.468910432914603,
+      "learning_rate": 4.113288718929254e-05,
+      "loss": 1.709,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3859408495366754,
+      "learning_rate": 4.1156787762906316e-05,
+      "loss": 0.7637,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.460804310963216,
+      "learning_rate": 4.118068833652008e-05,
+      "loss": 1.4766,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.186992758344865,
+      "learning_rate": 4.120458891013385e-05,
+      "loss": 1.127,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.834234352921078,
+      "learning_rate": 4.1228489483747614e-05,
+      "loss": 0.8589,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.943790809436944,
+      "learning_rate": 4.125239005736138e-05,
+      "loss": 1.6787,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5059062986501153,
+      "learning_rate": 4.127629063097515e-05,
+      "loss": 1.3008,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.455832852692391,
+      "learning_rate": 4.130019120458891e-05,
+      "loss": 1.2734,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.277668333291818,
+      "learning_rate": 4.132409177820268e-05,
+      "loss": 1.252,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.617692039025129,
+      "learning_rate": 4.1347992351816445e-05,
+      "loss": 0.7837,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7014448644422373,
+      "learning_rate": 4.137189292543021e-05,
+      "loss": 1.3398,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.420551395683192,
+      "learning_rate": 4.139579349904398e-05,
+      "loss": 0.9082,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.200659605118083,
+      "learning_rate": 4.1419694072657744e-05,
+      "loss": 0.6357,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.565048160361112,
+      "learning_rate": 4.144359464627151e-05,
+      "loss": 1.249,
+      "step": 1734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.718797493478208,
+      "learning_rate": 4.1467495219885276e-05,
+      "loss": 1.1699,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.894400116167546,
+      "learning_rate": 4.149139579349904e-05,
+      "loss": 1.374,
+      "step": 1736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1458799514516302,
+      "learning_rate": 4.151529636711281e-05,
+      "loss": 1.3311,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.97155645681188,
+      "learning_rate": 4.1539196940726575e-05,
+      "loss": 1.377,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.560557454857773,
+      "learning_rate": 4.156309751434035e-05,
+      "loss": 0.8027,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.038983873675596,
+      "learning_rate": 4.1586998087954114e-05,
+      "loss": 0.8057,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.576175637586296,
+      "learning_rate": 4.161089866156788e-05,
+      "loss": 1.1016,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.670187956932851,
+      "learning_rate": 4.1634799235181646e-05,
+      "loss": 1.8516,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.224325381260043,
+      "learning_rate": 4.165869980879541e-05,
+      "loss": 1.7656,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2463385141772108,
+      "learning_rate": 4.168260038240918e-05,
+      "loss": 0.9948,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.062227404526707,
+      "learning_rate": 4.170650095602295e-05,
+      "loss": 1.4355,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7081400484414866,
+      "learning_rate": 4.173040152963672e-05,
+      "loss": 1.4941,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8850193314826185,
+      "learning_rate": 4.1754302103250484e-05,
+      "loss": 0.9805,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.632088664507222,
+      "learning_rate": 4.177820267686425e-05,
+      "loss": 1.4141,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.655740038219902,
+      "learning_rate": 4.1802103250478016e-05,
+      "loss": 1.375,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.337976490760781,
+      "learning_rate": 4.182600382409178e-05,
+      "loss": 1.1602,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.339346870112514,
+      "learning_rate": 4.184990439770555e-05,
+      "loss": 0.8574,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.532416362055786,
+      "learning_rate": 4.1873804971319314e-05,
+      "loss": 1.1738,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9336745331025247,
+      "learning_rate": 4.189770554493308e-05,
+      "loss": 1.0117,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.360782329880484,
+      "learning_rate": 4.192160611854685e-05,
+      "loss": 1.0488,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4422192005875027,
+      "learning_rate": 4.194550669216061e-05,
+      "loss": 1.4824,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.683047514754976,
+      "learning_rate": 4.196940726577438e-05,
+      "loss": 1.6074,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.051389554735357,
+      "learning_rate": 4.1993307839388145e-05,
+      "loss": 1.3281,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8657590545615084,
+      "learning_rate": 4.201720841300191e-05,
+      "loss": 1.041,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4669525749741585,
+      "learning_rate": 4.204110898661568e-05,
+      "loss": 1.2207,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7305322293173133,
+      "learning_rate": 4.2065009560229444e-05,
+      "loss": 0.6396,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.332744504091536,
+      "learning_rate": 4.208891013384321e-05,
+      "loss": 1.3633,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.731821931446283,
+      "learning_rate": 4.2112810707456976e-05,
+      "loss": 0.6108,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.624167911889186,
+      "learning_rate": 4.213671128107074e-05,
+      "loss": 1.334,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.081357209688601,
+      "learning_rate": 4.2160611854684515e-05,
+      "loss": 0.9189,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.984131328760931,
+      "learning_rate": 4.218451242829828e-05,
+      "loss": 1.5312,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.262443489374537,
+      "learning_rate": 4.220841300191205e-05,
+      "loss": 0.9922,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.44918771899598,
+      "learning_rate": 4.2232313575525814e-05,
+      "loss": 1.8584,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.7484618256307645,
+      "learning_rate": 4.225621414913959e-05,
+      "loss": 0.7942,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1866107768188963,
+      "learning_rate": 4.228011472275335e-05,
+      "loss": 1.3105,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.232472325418228,
+      "learning_rate": 4.230401529636712e-05,
+      "loss": 1.1436,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.3796767359616755,
+      "learning_rate": 4.2327915869980885e-05,
+      "loss": 1.4785,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.501116725951554,
+      "learning_rate": 4.235181644359465e-05,
+      "loss": 1.3574,
+      "step": 1772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8025046011812957,
+      "learning_rate": 4.237571701720842e-05,
+      "loss": 1.209,
+      "step": 1773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.542815713445048,
+      "learning_rate": 4.2399617590822184e-05,
+      "loss": 1.2227,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.819287311912895,
+      "learning_rate": 4.242351816443595e-05,
+      "loss": 1.1406,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0971443744297766,
+      "learning_rate": 4.2447418738049716e-05,
+      "loss": 0.7437,
+      "step": 1776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9679483264985471,
+      "learning_rate": 4.247131931166348e-05,
+      "loss": 1.1631,
+      "step": 1777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.115419426566859,
+      "learning_rate": 4.249521988527725e-05,
+      "loss": 1.1973,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.462351412512986,
+      "learning_rate": 4.2519120458891015e-05,
+      "loss": 1.1934,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9991005840009866,
+      "learning_rate": 4.254302103250478e-05,
+      "loss": 0.9629,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.513633650672122,
+      "learning_rate": 4.256692160611855e-05,
+      "loss": 1.4512,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.463750643680711,
+      "learning_rate": 4.259082217973231e-05,
+      "loss": 0.7026,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.604496433421955,
+      "learning_rate": 4.261472275334608e-05,
+      "loss": 1.4258,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.885849798834554,
+      "learning_rate": 4.2638623326959845e-05,
+      "loss": 1.335,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2314367539173454,
+      "learning_rate": 4.266252390057361e-05,
+      "loss": 1.4053,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.190848013083002,
+      "learning_rate": 4.268642447418738e-05,
+      "loss": 1.3369,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9186415104218704,
+      "learning_rate": 4.271032504780115e-05,
+      "loss": 1.3262,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.4777836769968475,
+      "learning_rate": 4.273422562141492e-05,
+      "loss": 1.3848,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.245809972022742,
+      "learning_rate": 4.275812619502868e-05,
+      "loss": 1.1211,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7548790027613963,
+      "learning_rate": 4.278202676864245e-05,
+      "loss": 1.4668,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.469143407008934,
+      "learning_rate": 4.2805927342256215e-05,
+      "loss": 1.5312,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.307053132800666,
+      "learning_rate": 4.282982791586999e-05,
+      "loss": 2.0137,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.861496602068558,
+      "learning_rate": 4.2853728489483755e-05,
+      "loss": 1.166,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.376942955581478,
+      "learning_rate": 4.287762906309752e-05,
+      "loss": 1.9668,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1741033534509664,
+      "learning_rate": 4.290152963671129e-05,
+      "loss": 1.6465,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0530167085141398,
+      "learning_rate": 4.292543021032505e-05,
+      "loss": 0.7231,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.096462487246062,
+      "learning_rate": 4.294933078393882e-05,
+      "loss": 0.9678,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.91207926867628,
+      "learning_rate": 4.2973231357552585e-05,
+      "loss": 0.8054,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.571100085333846,
+      "learning_rate": 4.299713193116635e-05,
+      "loss": 1.2812,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6151531454723034,
+      "learning_rate": 4.302103250478012e-05,
+      "loss": 1.0596,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.432654730887218,
+      "learning_rate": 4.3044933078393884e-05,
+      "loss": 1.2031,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.589266586128263,
+      "learning_rate": 4.306883365200765e-05,
+      "loss": 1.1992,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.492216720288127,
+      "learning_rate": 4.3092734225621416e-05,
+      "loss": 0.9014,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.897020512848964,
+      "learning_rate": 4.311663479923518e-05,
+      "loss": 1.7109,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9644323171583284,
+      "learning_rate": 4.314053537284895e-05,
+      "loss": 1.4043,
+      "step": 1805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.553181100620422,
+      "learning_rate": 4.3164435946462715e-05,
+      "loss": 1.2383,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.441810390254913,
+      "learning_rate": 4.318833652007648e-05,
+      "loss": 1.752,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.604143117162401,
+      "learning_rate": 4.321223709369025e-05,
+      "loss": 0.9287,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.634319296464914,
+      "learning_rate": 4.323613766730401e-05,
+      "loss": 1.4297,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0614114500550755,
+      "learning_rate": 4.3260038240917786e-05,
+      "loss": 1.335,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.038184414251341,
+      "learning_rate": 4.328393881453155e-05,
+      "loss": 1.5645,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.316847863646837,
+      "learning_rate": 4.330783938814532e-05,
+      "loss": 0.8281,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.32626953014781,
+      "learning_rate": 4.3331739961759085e-05,
+      "loss": 0.7715,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.101731149234464,
+      "learning_rate": 4.335564053537285e-05,
+      "loss": 0.7769,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.061123424000534,
+      "learning_rate": 4.337954110898662e-05,
+      "loss": 1.0059,
+      "step": 1815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9849459369969593,
+      "learning_rate": 4.340344168260038e-05,
+      "loss": 1.6113,
+      "step": 1816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1134535527621927,
+      "learning_rate": 4.342734225621415e-05,
+      "loss": 0.9434,
+      "step": 1817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7389028499403065,
+      "learning_rate": 4.345124282982792e-05,
+      "loss": 0.9434,
+      "step": 1818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.535105935465898,
+      "learning_rate": 4.347514340344169e-05,
+      "loss": 0.9805,
+      "step": 1819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2565852719316006,
+      "learning_rate": 4.3499043977055455e-05,
+      "loss": 1.0791,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.812512207004759,
+      "learning_rate": 4.352294455066922e-05,
+      "loss": 0.9019,
+      "step": 1821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7102127205669033,
+      "learning_rate": 4.354684512428299e-05,
+      "loss": 1.1455,
+      "step": 1822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.43561196708016,
+      "learning_rate": 4.357074569789675e-05,
+      "loss": 0.7363,
+      "step": 1823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.078974590723116,
+      "learning_rate": 4.359464627151052e-05,
+      "loss": 1.8496,
+      "step": 1824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.546024110162991,
+      "learning_rate": 4.3618546845124286e-05,
+      "loss": 1.2969,
+      "step": 1825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.625121899449557,
+      "learning_rate": 4.364244741873805e-05,
+      "loss": 0.8838,
+      "step": 1826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8564686816761555,
+      "learning_rate": 4.366634799235182e-05,
+      "loss": 1.4434,
+      "step": 1827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9241666731603853,
+      "learning_rate": 4.3690248565965584e-05,
+      "loss": 1.2012,
+      "step": 1828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7026749349475927,
+      "learning_rate": 4.371414913957935e-05,
+      "loss": 1.4121,
+      "step": 1829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.232821809597205,
+      "learning_rate": 4.3738049713193116e-05,
+      "loss": 1.2031,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5927954691496273,
+      "learning_rate": 4.376195028680688e-05,
+      "loss": 1.4043,
+      "step": 1831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4796606474626133,
+      "learning_rate": 4.378585086042065e-05,
+      "loss": 1.3457,
+      "step": 1832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.830869136037251,
+      "learning_rate": 4.380975143403442e-05,
+      "loss": 1.5078,
+      "step": 1833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.318443723867277,
+      "learning_rate": 4.383365200764819e-05,
+      "loss": 1.0625,
+      "step": 1834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.282866963096516,
+      "learning_rate": 4.3857552581261954e-05,
+      "loss": 1.0859,
+      "step": 1835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8480195911247983,
+      "learning_rate": 4.388145315487572e-05,
+      "loss": 1.7637,
+      "step": 1836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.61024590095925,
+      "learning_rate": 4.3905353728489486e-05,
+      "loss": 0.8027,
+      "step": 1837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.282985586059741,
+      "learning_rate": 4.392925430210325e-05,
+      "loss": 0.8545,
+      "step": 1838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.099078589064222,
+      "learning_rate": 4.395315487571702e-05,
+      "loss": 1.3672,
+      "step": 1839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.264038571410627,
+      "learning_rate": 4.3977055449330785e-05,
+      "loss": 1.5154,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.551296865508475,
+      "learning_rate": 4.400095602294455e-05,
+      "loss": 1.0527,
+      "step": 1841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.1606346046238984,
+      "learning_rate": 4.402485659655832e-05,
+      "loss": 2.5527,
+      "step": 1842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.462292959034473,
+      "learning_rate": 4.4048757170172083e-05,
+      "loss": 0.791,
+      "step": 1843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9619353653220113,
+      "learning_rate": 4.4072657743785856e-05,
+      "loss": 1.0156,
+      "step": 1844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9777353934534365,
+      "learning_rate": 4.409655831739962e-05,
+      "loss": 0.8223,
+      "step": 1845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.291315650782221,
+      "learning_rate": 4.412045889101339e-05,
+      "loss": 1.3047,
+      "step": 1846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0988407059258667,
+      "learning_rate": 4.4144359464627155e-05,
+      "loss": 1.0269,
+      "step": 1847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.515678081311061,
+      "learning_rate": 4.416826003824092e-05,
+      "loss": 1.9512,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1420905646002475,
+      "learning_rate": 4.419216061185469e-05,
+      "loss": 0.7588,
+      "step": 1849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5330247661827094,
+      "learning_rate": 4.4216061185468453e-05,
+      "loss": 1.625,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.471577359822819,
+      "learning_rate": 4.423996175908222e-05,
+      "loss": 1.0938,
+      "step": 1851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.52849351420399,
+      "learning_rate": 4.4263862332695986e-05,
+      "loss": 1.123,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3287533929690563,
+      "learning_rate": 4.428776290630975e-05,
+      "loss": 0.8623,
+      "step": 1853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.640100795026777,
+      "learning_rate": 4.431166347992352e-05,
+      "loss": 0.8701,
+      "step": 1854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.955688885640272,
+      "learning_rate": 4.4335564053537284e-05,
+      "loss": 1.7266,
+      "step": 1855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6279603523271273,
+      "learning_rate": 4.435946462715106e-05,
+      "loss": 1.2188,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7888410036840297,
+      "learning_rate": 4.4383365200764823e-05,
+      "loss": 0.9102,
+      "step": 1857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7003696930299053,
+      "learning_rate": 4.440726577437859e-05,
+      "loss": 0.9971,
+      "step": 1858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2737421310146035,
+      "learning_rate": 4.4431166347992356e-05,
+      "loss": 0.9844,
+      "step": 1859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.151391932037589,
+      "learning_rate": 4.445506692160612e-05,
+      "loss": 1.4976,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.174322527598024,
+      "learning_rate": 4.447896749521989e-05,
+      "loss": 1.4062,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2994517709576883,
+      "learning_rate": 4.4502868068833654e-05,
+      "loss": 0.9033,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.131456999360886,
+      "learning_rate": 4.452676864244742e-05,
+      "loss": 1.2441,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.438373385362679,
+      "learning_rate": 4.455066921606119e-05,
+      "loss": 1.3105,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.621830070756279,
+      "learning_rate": 4.457456978967495e-05,
+      "loss": 0.9854,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.086493942964978,
+      "learning_rate": 4.459847036328872e-05,
+      "loss": 1.1367,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4078683736508273,
+      "learning_rate": 4.4622370936902485e-05,
+      "loss": 1.1465,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3337838504247,
+      "learning_rate": 4.464627151051625e-05,
+      "loss": 1.3496,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.714657694939116,
+      "learning_rate": 4.467017208413002e-05,
+      "loss": 1.0146,
+      "step": 1869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.0931195837770185,
+      "learning_rate": 4.469407265774379e-05,
+      "loss": 1.4648,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.747428982674659,
+      "learning_rate": 4.4717973231357557e-05,
+      "loss": 1.3066,
+      "step": 1871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.009099690754224,
+      "learning_rate": 4.474187380497132e-05,
+      "loss": 0.8311,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3756371355889234,
+      "learning_rate": 4.476577437858509e-05,
+      "loss": 1.3008,
+      "step": 1873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.797758063876688,
+      "learning_rate": 4.4789674952198855e-05,
+      "loss": 1.5566,
+      "step": 1874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.568224545169976,
+      "learning_rate": 4.481357552581262e-05,
+      "loss": 0.9189,
+      "step": 1875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6937992879159838,
+      "learning_rate": 4.483747609942639e-05,
+      "loss": 1.3477,
+      "step": 1876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.8156746019219,
+      "learning_rate": 4.4861376673040154e-05,
+      "loss": 1.6934,
+      "step": 1877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5801729563272215,
+      "learning_rate": 4.488527724665392e-05,
+      "loss": 1.208,
+      "step": 1878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6337246313741596,
+      "learning_rate": 4.490917782026769e-05,
+      "loss": 1.1416,
+      "step": 1879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5598675555334154,
+      "learning_rate": 4.493307839388146e-05,
+      "loss": 0.8613,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8159735443109857,
+      "learning_rate": 4.4956978967495225e-05,
+      "loss": 0.7432,
+      "step": 1881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8861317867092304,
+      "learning_rate": 4.498087954110899e-05,
+      "loss": 1.0869,
+      "step": 1882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.841772671862309,
+      "learning_rate": 4.500478011472276e-05,
+      "loss": 0.7588,
+      "step": 1883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.364273140568858,
+      "learning_rate": 4.5028680688336524e-05,
+      "loss": 1.4238,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2335056481867275,
+      "learning_rate": 4.505258126195029e-05,
+      "loss": 0.9756,
+      "step": 1885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.268064030240045,
+      "learning_rate": 4.5076481835564056e-05,
+      "loss": 0.6733,
+      "step": 1886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.127094777280887,
+      "learning_rate": 4.510038240917782e-05,
+      "loss": 0.8604,
+      "step": 1887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5956525144673344,
+      "learning_rate": 4.512428298279159e-05,
+      "loss": 0.7549,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.540366217005131,
+      "learning_rate": 4.5148183556405354e-05,
+      "loss": 1.5703,
+      "step": 1889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6862738717688392,
+      "learning_rate": 4.517208413001912e-05,
+      "loss": 0.873,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.249633017387757,
+      "learning_rate": 4.519598470363289e-05,
+      "loss": 0.7378,
+      "step": 1891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.658991678538754,
+      "learning_rate": 4.521988527724665e-05,
+      "loss": 0.5287,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7282590643068807,
+      "learning_rate": 4.524378585086042e-05,
+      "loss": 0.9575,
+      "step": 1893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9304370119518657,
+      "learning_rate": 4.5267686424474185e-05,
+      "loss": 0.894,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1452323558468023,
+      "learning_rate": 4.529158699808795e-05,
+      "loss": 1.3262,
+      "step": 1895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5803122685810185,
+      "learning_rate": 4.5315487571701724e-05,
+      "loss": 0.9209,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4611933499506016,
+      "learning_rate": 4.533938814531549e-05,
+      "loss": 0.9365,
+      "step": 1897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.982633920873581,
+      "learning_rate": 4.536328871892926e-05,
+      "loss": 1.1504,
+      "step": 1898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4512526054310118,
+      "learning_rate": 4.538718929254302e-05,
+      "loss": 1.1121,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1748706823030215,
+      "learning_rate": 4.541108986615679e-05,
+      "loss": 0.8589,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8548564959000333,
+      "learning_rate": 4.5434990439770555e-05,
+      "loss": 1.2158,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.196709437105719,
+      "learning_rate": 4.545889101338433e-05,
+      "loss": 0.7676,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6759196224771498,
+      "learning_rate": 4.5482791586998094e-05,
+      "loss": 1.4551,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8168259658731751,
+      "learning_rate": 4.550669216061186e-05,
+      "loss": 1.2344,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.8745748872512875,
+      "learning_rate": 4.553059273422563e-05,
+      "loss": 1.1758,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.163112770996092,
+      "learning_rate": 4.555449330783939e-05,
+      "loss": 1.4082,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3452645552725997,
+      "learning_rate": 4.557839388145316e-05,
+      "loss": 1.0,
+      "step": 1907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7180559808991056,
+      "learning_rate": 4.5602294455066925e-05,
+      "loss": 0.8213,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0952005628118235,
+      "learning_rate": 4.562619502868069e-05,
+      "loss": 0.9883,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6963567065548197,
+      "learning_rate": 4.565009560229446e-05,
+      "loss": 0.752,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.306160409123862,
+      "learning_rate": 4.5673996175908224e-05,
+      "loss": 1.2676,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.477953274884884,
+      "learning_rate": 4.569789674952199e-05,
+      "loss": 2.2637,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.488678665403348,
+      "learning_rate": 4.5721797323135756e-05,
+      "loss": 0.9229,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.397468092508916,
+      "learning_rate": 4.574569789674952e-05,
+      "loss": 1.4551,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.509481094796855,
+      "learning_rate": 4.576959847036329e-05,
+      "loss": 0.7168,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.157617962951393,
+      "learning_rate": 4.5793499043977055e-05,
+      "loss": 0.7271,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.74781027224713,
+      "learning_rate": 4.581739961759082e-05,
+      "loss": 0.8047,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.397998158787256,
+      "learning_rate": 4.584130019120459e-05,
+      "loss": 0.9312,
+      "step": 1918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6316657129841245,
+      "learning_rate": 4.586520076481835e-05,
+      "loss": 0.814,
+      "step": 1919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.136621560182516,
+      "learning_rate": 4.588910133843212e-05,
+      "loss": 1.2,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.257233914602026,
+      "learning_rate": 4.5913001912045885e-05,
+      "loss": 1.0869,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8184750983258327,
+      "learning_rate": 4.593690248565966e-05,
+      "loss": 0.8633,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.319225465440648,
+      "learning_rate": 4.5960803059273425e-05,
+      "loss": 1.1699,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8817123604543653,
+      "learning_rate": 4.598470363288719e-05,
+      "loss": 1.5195,
+      "step": 1924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.705047181892874,
+      "learning_rate": 4.6008604206500964e-05,
+      "loss": 1.4082,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6803951830373363,
+      "learning_rate": 4.603250478011473e-05,
+      "loss": 1.125,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.197688490142887,
+      "learning_rate": 4.6056405353728496e-05,
+      "loss": 1.3125,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.808897775789497,
+      "learning_rate": 4.608030592734226e-05,
+      "loss": 1.7051,
+      "step": 1928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6921660260247933,
+      "learning_rate": 4.610420650095603e-05,
+      "loss": 1.9609,
+      "step": 1929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2561397493852424,
+      "learning_rate": 4.6128107074569795e-05,
+      "loss": 1.377,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0431511448465485,
+      "learning_rate": 4.615200764818356e-05,
+      "loss": 1.748,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.176604575731536,
+      "learning_rate": 4.617590822179733e-05,
+      "loss": 1.166,
+      "step": 1932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7446258182788967,
+      "learning_rate": 4.619980879541109e-05,
+      "loss": 0.8604,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.030851170624764,
+      "learning_rate": 4.622370936902486e-05,
+      "loss": 0.7915,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8668040075794154,
+      "learning_rate": 4.6247609942638625e-05,
+      "loss": 1.7812,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.730366746396669,
+      "learning_rate": 4.627151051625239e-05,
+      "loss": 1.4004,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.431715082423208,
+      "learning_rate": 4.629541108986616e-05,
+      "loss": 1.0977,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0930677982144688,
+      "learning_rate": 4.6319311663479924e-05,
+      "loss": 1.3203,
+      "step": 1938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8790160614975457,
+      "learning_rate": 4.634321223709369e-05,
+      "loss": 1.0762,
+      "step": 1939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.663727970362655,
+      "learning_rate": 4.6367112810707456e-05,
+      "loss": 1.2324,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.169541726907213,
+      "learning_rate": 4.639101338432122e-05,
+      "loss": 0.5017,
+      "step": 1941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.252534180952343,
+      "learning_rate": 4.641491395793499e-05,
+      "loss": 0.6426,
+      "step": 1942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.63213855685818,
+      "learning_rate": 4.6438814531548755e-05,
+      "loss": 0.9932,
+      "step": 1943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.00016345710601,
+      "learning_rate": 4.646271510516252e-05,
+      "loss": 1.0967,
+      "step": 1944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.90192898519337,
+      "learning_rate": 4.6486615678776294e-05,
+      "loss": 1.1465,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.347003185101683,
+      "learning_rate": 4.651051625239006e-05,
+      "loss": 0.8848,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.455681746152663,
+      "learning_rate": 4.6534416826003826e-05,
+      "loss": 0.7998,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.540490602736801,
+      "learning_rate": 4.655831739961759e-05,
+      "loss": 0.8813,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.595762117555253,
+      "learning_rate": 4.6582217973231365e-05,
+      "loss": 0.8965,
+      "step": 1949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.520145595305704,
+      "learning_rate": 4.660611854684513e-05,
+      "loss": 1.7656,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0211873515731633,
+      "learning_rate": 4.66300191204589e-05,
+      "loss": 1.0195,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.155274909393168,
+      "learning_rate": 4.6653919694072664e-05,
+      "loss": 0.9395,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.220417512606175,
+      "learning_rate": 4.667782026768643e-05,
+      "loss": 0.8452,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.586660141944859,
+      "learning_rate": 4.6701720841300196e-05,
+      "loss": 0.9512,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2537642885454447,
+      "learning_rate": 4.672562141491396e-05,
+      "loss": 1.3086,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.606027174613974,
+      "learning_rate": 4.674952198852773e-05,
+      "loss": 1.167,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.095700974573029,
+      "learning_rate": 4.6773422562141495e-05,
+      "loss": 1.4375,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.07888992868052,
+      "learning_rate": 4.679732313575526e-05,
+      "loss": 1.8896,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8598909173183444,
+      "learning_rate": 4.682122370936903e-05,
+      "loss": 1.1152,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.001496273566697,
+      "learning_rate": 4.684512428298279e-05,
+      "loss": 1.0791,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0089640310857195,
+      "learning_rate": 4.686902485659656e-05,
+      "loss": 0.96,
+      "step": 1961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.201897461793118,
+      "learning_rate": 4.6892925430210326e-05,
+      "loss": 0.8477,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.010699743543679,
+      "learning_rate": 4.691682600382409e-05,
+      "loss": 1.3184,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6871967588163805,
+      "learning_rate": 4.694072657743786e-05,
+      "loss": 1.0439,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.534017204117877,
+      "learning_rate": 4.6964627151051624e-05,
+      "loss": 1.3789,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.687614438481473,
+      "learning_rate": 4.698852772466539e-05,
+      "loss": 1.0947,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.81904833286989,
+      "learning_rate": 4.7012428298279156e-05,
+      "loss": 1.0928,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7327491769322245,
+      "learning_rate": 4.703632887189293e-05,
+      "loss": 0.7876,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9691873700375568,
+      "learning_rate": 4.7060229445506696e-05,
+      "loss": 1.2021,
+      "step": 1969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.029693363766699,
+      "learning_rate": 4.708413001912046e-05,
+      "loss": 1.7559,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.988626856546481,
+      "learning_rate": 4.710803059273423e-05,
+      "loss": 1.5352,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.910740351382798,
+      "learning_rate": 4.7131931166347994e-05,
+      "loss": 1.2334,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.820816206398776,
+      "learning_rate": 4.715583173996176e-05,
+      "loss": 0.9766,
+      "step": 1973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.529822429712984,
+      "learning_rate": 4.7179732313575526e-05,
+      "loss": 1.0918,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5939463686160207,
+      "learning_rate": 4.72036328871893e-05,
+      "loss": 1.4814,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.4670958714903115,
+      "learning_rate": 4.7227533460803066e-05,
+      "loss": 0.7646,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.526683048457826,
+      "learning_rate": 4.725143403441683e-05,
+      "loss": 1.2852,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9061896205085382,
+      "learning_rate": 4.72753346080306e-05,
+      "loss": 0.7998,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.257049772208627,
+      "learning_rate": 4.7299235181644364e-05,
+      "loss": 0.9463,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.534113219401744,
+      "learning_rate": 4.732313575525813e-05,
+      "loss": 2.1621,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.180436967341457,
+      "learning_rate": 4.7347036328871896e-05,
+      "loss": 0.749,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6048287630819895,
+      "learning_rate": 4.737093690248566e-05,
+      "loss": 1.2207,
+      "step": 1982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.522634613518883,
+      "learning_rate": 4.739483747609943e-05,
+      "loss": 1.2617,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.909533190378599,
+      "learning_rate": 4.7418738049713195e-05,
+      "loss": 1.248,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5645207136882866,
+      "learning_rate": 4.744263862332696e-05,
+      "loss": 1.4238,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.534752106744418,
+      "learning_rate": 4.746653919694073e-05,
+      "loss": 1.0723,
+      "step": 1986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.083248773051081,
+      "learning_rate": 4.7490439770554493e-05,
+      "loss": 0.9863,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.692795470766261,
+      "learning_rate": 4.751434034416826e-05,
+      "loss": 0.9092,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.181889669193641,
+      "learning_rate": 4.7538240917782026e-05,
+      "loss": 1.6035,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.350133112931805,
+      "learning_rate": 4.756214149139579e-05,
+      "loss": 0.6409,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.688139284970027,
+      "learning_rate": 4.7586042065009565e-05,
+      "loss": 0.7988,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2700678967308052,
+      "learning_rate": 4.760994263862333e-05,
+      "loss": 1.2383,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6258663589164684,
+      "learning_rate": 4.76338432122371e-05,
+      "loss": 1.9141,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8454375237989784,
+      "learning_rate": 4.7657743785850863e-05,
+      "loss": 0.7073,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.896599481033867,
+      "learning_rate": 4.768164435946463e-05,
+      "loss": 1.5801,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6577326507504457,
+      "learning_rate": 4.7705544933078396e-05,
+      "loss": 1.375,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.197120354624175,
+      "learning_rate": 4.772944550669216e-05,
+      "loss": 1.6367,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.838673685341397,
+      "learning_rate": 4.775334608030593e-05,
+      "loss": 2.0273,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.695019197109898,
+      "learning_rate": 4.7777246653919694e-05,
+      "loss": 1.3535,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8795442085367724,
+      "learning_rate": 4.780114722753346e-05,
+      "loss": 1.3398,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.41136042902183,
+      "learning_rate": 4.782504780114723e-05,
+      "loss": 1.2256,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0968535298393305,
+      "learning_rate": 4.7848948374761e-05,
+      "loss": 1.1758,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.849055691659198,
+      "learning_rate": 4.7872848948374766e-05,
+      "loss": 1.543,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.712472744566925,
+      "learning_rate": 4.789674952198853e-05,
+      "loss": 0.9814,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.887733952015752,
+      "learning_rate": 4.79206500956023e-05,
+      "loss": 1.8633,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.53872017710906,
+      "learning_rate": 4.7944550669216064e-05,
+      "loss": 2.4883,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.083305155245637,
+      "learning_rate": 4.796845124282983e-05,
+      "loss": 1.1416,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9246168544876845,
+      "learning_rate": 4.7992351816443597e-05,
+      "loss": 1.0449,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1796660610224774,
+      "learning_rate": 4.801625239005736e-05,
+      "loss": 1.4453,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1836975455184704,
+      "learning_rate": 4.804015296367113e-05,
+      "loss": 1.2891,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2636590254569513,
+      "learning_rate": 4.8064053537284895e-05,
+      "loss": 0.999,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9115314637656704,
+      "learning_rate": 4.808795411089866e-05,
+      "loss": 1.3467,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.202692972892242,
+      "learning_rate": 4.811185468451243e-05,
+      "loss": 0.8643,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.695101787789361,
+      "learning_rate": 4.81357552581262e-05,
+      "loss": 1.3164,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2715017985547363,
+      "learning_rate": 4.8159655831739967e-05,
+      "loss": 1.0723,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.115360386582984,
+      "learning_rate": 4.818355640535373e-05,
+      "loss": 0.4978,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.355039319647087,
+      "learning_rate": 4.82074569789675e-05,
+      "loss": 1.4043,
+      "step": 2017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.696192843045433,
+      "learning_rate": 4.8231357552581265e-05,
+      "loss": 1.2461,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5456304961551908,
+      "learning_rate": 4.825525812619503e-05,
+      "loss": 1.0137,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.769122331915219,
+      "learning_rate": 4.82791586998088e-05,
+      "loss": 1.1211,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.711522510447666,
+      "learning_rate": 4.8303059273422564e-05,
+      "loss": 0.8955,
+      "step": 2021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.76414529016536,
+      "learning_rate": 4.832695984703633e-05,
+      "loss": 1.0663,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3847535279113226,
+      "learning_rate": 4.8350860420650096e-05,
+      "loss": 1.6797,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9975785259333105,
+      "learning_rate": 4.837476099426386e-05,
+      "loss": 1.1465,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4873046875,
+      "learning_rate": 4.839866156787763e-05,
+      "loss": 1.085,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.058304741051677,
+      "learning_rate": 4.8422562141491394e-05,
+      "loss": 0.9741,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.71650059042211,
+      "learning_rate": 4.844646271510517e-05,
+      "loss": 1.4766,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.281844312342391,
+      "learning_rate": 4.8470363288718934e-05,
+      "loss": 1.0984,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.437909752110413,
+      "learning_rate": 4.84942638623327e-05,
+      "loss": 1.1152,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0762289475413453,
+      "learning_rate": 4.8518164435946466e-05,
+      "loss": 1.334,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.576878194166263,
+      "learning_rate": 4.854206500956023e-05,
+      "loss": 1.0093,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0151634060686345,
+      "learning_rate": 4.8565965583174e-05,
+      "loss": 1.2363,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.496481613088413,
+      "learning_rate": 4.8589866156787764e-05,
+      "loss": 1.0247,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.662247204308677,
+      "learning_rate": 4.861376673040153e-05,
+      "loss": 0.5632,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.1273970141810805,
+      "learning_rate": 4.86376673040153e-05,
+      "loss": 0.9756,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4763903628820456,
+      "learning_rate": 4.866156787762906e-05,
+      "loss": 1.127,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.68052042036029,
+      "learning_rate": 4.8685468451242836e-05,
+      "loss": 0.9688,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.052550209625977,
+      "learning_rate": 4.87093690248566e-05,
+      "loss": 1.5312,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.171370937189302,
+      "learning_rate": 4.873326959847037e-05,
+      "loss": 1.7246,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5411106944801265,
+      "learning_rate": 4.8757170172084134e-05,
+      "loss": 0.9419,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6219428753026417,
+      "learning_rate": 4.87810707456979e-05,
+      "loss": 0.7161,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0167069008126592,
+      "learning_rate": 4.880497131931167e-05,
+      "loss": 0.8569,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.756753942925076,
+      "learning_rate": 4.882887189292543e-05,
+      "loss": 1.1426,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8364902572001025,
+      "learning_rate": 4.88527724665392e-05,
+      "loss": 1.2002,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.976592616932612,
+      "learning_rate": 4.8876673040152965e-05,
+      "loss": 1.5801,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.920733060398951,
+      "learning_rate": 4.890057361376673e-05,
+      "loss": 0.9619,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.10584432617426,
+      "learning_rate": 4.89244741873805e-05,
+      "loss": 1.2539,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.993004164802716,
+      "learning_rate": 4.8948374760994264e-05,
+      "loss": 1.5527,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.756054551640807,
+      "learning_rate": 4.897227533460803e-05,
+      "loss": 1.6914,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2429795411427325,
+      "learning_rate": 4.8996175908221796e-05,
+      "loss": 0.616,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.377238871102795,
+      "learning_rate": 4.902007648183556e-05,
+      "loss": 1.9062,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.179276031352178,
+      "learning_rate": 4.904397705544933e-05,
+      "loss": 1.584,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.858712062212828,
+      "learning_rate": 4.90678776290631e-05,
+      "loss": 0.8945,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.310146629512215,
+      "learning_rate": 4.909177820267687e-05,
+      "loss": 1.2666,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.959729473451669,
+      "learning_rate": 4.9115678776290634e-05,
+      "loss": 1.7031,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.15988419513719,
+      "learning_rate": 4.91395793499044e-05,
+      "loss": 1.7363,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1521835570066328,
+      "learning_rate": 4.9163479923518166e-05,
+      "loss": 0.9521,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.393694739264854,
+      "learning_rate": 4.918738049713193e-05,
+      "loss": 1.1816,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0179848406128458,
+      "learning_rate": 4.92112810707457e-05,
+      "loss": 1.4082,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.1904075471364575,
+      "learning_rate": 4.923518164435947e-05,
+      "loss": 1.2715,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.184084988940602,
+      "learning_rate": 4.925908221797324e-05,
+      "loss": 1.7344,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.018939479792689,
+      "learning_rate": 4.9282982791587004e-05,
+      "loss": 1.1709,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.25505440231024,
+      "learning_rate": 4.930688336520077e-05,
+      "loss": 1.3105,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5463290025357312,
+      "learning_rate": 4.9330783938814536e-05,
+      "loss": 1.1113,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.401641166831737,
+      "learning_rate": 4.93546845124283e-05,
+      "loss": 0.7627,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.133236388117557,
+      "learning_rate": 4.937858508604207e-05,
+      "loss": 2.1914,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.313769430267508,
+      "learning_rate": 4.9402485659655835e-05,
+      "loss": 0.5598,
+      "step": 2067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.337091591998113,
+      "learning_rate": 4.94263862332696e-05,
+      "loss": 1.1562,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.503414308047369,
+      "learning_rate": 4.945028680688337e-05,
+      "loss": 0.96,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.482956775388913,
+      "learning_rate": 4.947418738049713e-05,
+      "loss": 1.2129,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3234258267624535,
+      "learning_rate": 4.94980879541109e-05,
+      "loss": 1.1162,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6159749880478675,
+      "learning_rate": 4.9521988527724665e-05,
+      "loss": 1.1172,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.50394817508132,
+      "learning_rate": 4.954588910133843e-05,
+      "loss": 0.8008,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6250538593398067,
+      "learning_rate": 4.95697896749522e-05,
+      "loss": 1.498,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.768500785115937,
+      "learning_rate": 4.9593690248565964e-05,
+      "loss": 0.9824,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.696695785490989,
+      "learning_rate": 4.961759082217973e-05,
+      "loss": 1.251,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.202528139656313,
+      "learning_rate": 4.9641491395793496e-05,
+      "loss": 1.1914,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.008153643397417,
+      "learning_rate": 4.966539196940726e-05,
+      "loss": 0.8496,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.811160467951751,
+      "learning_rate": 4.9689292543021035e-05,
+      "loss": 0.9658,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.582093577130455,
+      "learning_rate": 4.97131931166348e-05,
+      "loss": 1.375,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8506663926483486,
+      "learning_rate": 4.973709369024857e-05,
+      "loss": 1.1138,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.084065404131002,
+      "learning_rate": 4.9760994263862334e-05,
+      "loss": 1.0664,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.950223322196851,
+      "learning_rate": 4.978489483747611e-05,
+      "loss": 1.1543,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.848029469337586,
+      "learning_rate": 4.980879541108987e-05,
+      "loss": 0.9629,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.437771426214203,
+      "learning_rate": 4.983269598470364e-05,
+      "loss": 0.6094,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3379947444768594,
+      "learning_rate": 4.9856596558317405e-05,
+      "loss": 1.3057,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.400217396108941,
+      "learning_rate": 4.988049713193117e-05,
+      "loss": 0.7871,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.363401803969358,
+      "learning_rate": 4.990439770554494e-05,
+      "loss": 1.4307,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7007268552545387,
+      "learning_rate": 4.9928298279158704e-05,
+      "loss": 1.0469,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.250545242339826,
+      "learning_rate": 4.995219885277247e-05,
+      "loss": 0.9727,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.758188589710484,
+      "learning_rate": 4.9976099426386236e-05,
+      "loss": 0.8945,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.766312594526509,
+      "learning_rate": 5e-05,
+      "loss": 1.0283,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8676951943754827,
+      "learning_rate": 4.999999965213009e-05,
+      "loss": 0.9902,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4051472502615456,
+      "learning_rate": 4.999999860852037e-05,
+      "loss": 0.8799,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1755503561074443,
+      "learning_rate": 4.999999686917086e-05,
+      "loss": 1.1299,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.628943367327581,
+      "learning_rate": 4.9999994434081616e-05,
+      "loss": 1.6152,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.117589633023256,
+      "learning_rate": 4.99999913032527e-05,
+      "loss": 1.1504,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.382310916806065,
+      "learning_rate": 4.999998747668422e-05,
+      "loss": 1.2188,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.135807156719295,
+      "learning_rate": 4.9999982954376247e-05,
+      "loss": 0.8694,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1562172160948205,
+      "learning_rate": 4.999997773632893e-05,
+      "loss": 1.5098,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2114255956481816,
+      "learning_rate": 4.9999971822542416e-05,
+      "loss": 1.5078,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.044461739202831,
+      "learning_rate": 4.999996521301687e-05,
+      "loss": 0.588,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.670753574059004,
+      "learning_rate": 4.999995790775246e-05,
+      "loss": 0.7256,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.315625813212722,
+      "learning_rate": 4.9999949906749396e-05,
+      "loss": 1.209,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.013944633916616,
+      "learning_rate": 4.999994121000791e-05,
+      "loss": 1.4551,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8284396001858365,
+      "learning_rate": 4.9999931817528235e-05,
+      "loss": 1.5273,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.921131400530967,
+      "learning_rate": 4.999992172931063e-05,
+      "loss": 1.7656,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9462990915255514,
+      "learning_rate": 4.999991094535539e-05,
+      "loss": 0.9229,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.983121018463694,
+      "learning_rate": 4.99998994656628e-05,
+      "loss": 1.1533,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.817368632080459,
+      "learning_rate": 4.999988729023318e-05,
+      "loss": 1.375,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.768072306272928,
+      "learning_rate": 4.999987441906689e-05,
+      "loss": 0.8633,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.373387543209944,
+      "learning_rate": 4.999986085216426e-05,
+      "loss": 2.1113,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0897120924504162,
+      "learning_rate": 4.999984658952568e-05,
+      "loss": 0.4878,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5475035283866307,
+      "learning_rate": 4.999983163115156e-05,
+      "loss": 1.1445,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.409955824767654,
+      "learning_rate": 4.999981597704229e-05,
+      "loss": 1.6992,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.853498548729904,
+      "learning_rate": 4.999979962719832e-05,
+      "loss": 0.9932,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.199804229106597,
+      "learning_rate": 4.9999782581620106e-05,
+      "loss": 0.4646,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.926811576708381,
+      "learning_rate": 4.999976484030812e-05,
+      "loss": 1.1855,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.142657615741619,
+      "learning_rate": 4.999974640326286e-05,
+      "loss": 0.9443,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.683759192716649,
+      "learning_rate": 4.999972727048483e-05,
+      "loss": 0.8774,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7937408447115613,
+      "learning_rate": 4.999970744197456e-05,
+      "loss": 0.9893,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.184579595969296,
+      "learning_rate": 4.999968691773262e-05,
+      "loss": 1.3613,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4691853199809,
+      "learning_rate": 4.999966569775957e-05,
+      "loss": 0.7627,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.765191530619344,
+      "learning_rate": 4.9999643782056005e-05,
+      "loss": 0.8657,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.617139939967072,
+      "learning_rate": 4.999962117062252e-05,
+      "loss": 1.9219,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.841710983652239,
+      "learning_rate": 4.999959786345976e-05,
+      "loss": 0.9033,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.672972069094517,
+      "learning_rate": 4.999957386056837e-05,
+      "loss": 1.3232,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.968974704284346,
+      "learning_rate": 4.999954916194902e-05,
+      "loss": 0.8867,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.003660093571192,
+      "learning_rate": 4.999952376760238e-05,
+      "loss": 0.7354,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5214185660968176,
+      "learning_rate": 4.9999497677529186e-05,
+      "loss": 1.6113,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.353448966574173,
+      "learning_rate": 4.999947089173015e-05,
+      "loss": 1.2852,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.136459418534763,
+      "learning_rate": 4.999944341020601e-05,
+      "loss": 1.1445,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3163390383047893,
+      "learning_rate": 4.9999415232957536e-05,
+      "loss": 0.9082,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9865904568701533,
+      "learning_rate": 4.999938635998551e-05,
+      "loss": 1.4297,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7797664157243087,
+      "learning_rate": 4.999935679129075e-05,
+      "loss": 2.1328,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.137749738771514,
+      "learning_rate": 4.9999326526874055e-05,
+      "loss": 0.8369,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.039115171565788,
+      "learning_rate": 4.9999295566736295e-05,
+      "loss": 1.5254,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5023640414417487,
+      "learning_rate": 4.99992639108783e-05,
+      "loss": 1.2212,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.955531426097764,
+      "learning_rate": 4.999923155930098e-05,
+      "loss": 1.4453,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.403467724870086,
+      "learning_rate": 4.999919851200522e-05,
+      "loss": 1.2754,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.566727135761939,
+      "learning_rate": 4.9999164768991935e-05,
+      "loss": 1.8281,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2673641501447386,
+      "learning_rate": 4.999913033026208e-05,
+      "loss": 1.5742,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.971596532248748,
+      "learning_rate": 4.99990951958166e-05,
+      "loss": 1.2012,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.213259819170069,
+      "learning_rate": 4.999905936565649e-05,
+      "loss": 1.0957,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.088814341058083,
+      "learning_rate": 4.9999022839782725e-05,
+      "loss": 1.3945,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4286694126237536,
+      "learning_rate": 4.999898561819632e-05,
+      "loss": 0.7163,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9696551097639716,
+      "learning_rate": 4.999894770089834e-05,
+      "loss": 0.793,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2670359506581903,
+      "learning_rate": 4.9998909087889824e-05,
+      "loss": 1.2451,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.4818875241124125,
+      "learning_rate": 4.9998869779171834e-05,
+      "loss": 1.269,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0404154847463287,
+      "learning_rate": 4.999882977474548e-05,
+      "loss": 1.5273,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.557134453835041,
+      "learning_rate": 4.999878907461187e-05,
+      "loss": 1.0889,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.681626161911848,
+      "learning_rate": 4.999874767877214e-05,
+      "loss": 0.8818,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8958171225361817,
+      "learning_rate": 4.999870558722743e-05,
+      "loss": 0.9355,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.375470340465622,
+      "learning_rate": 4.999866279997892e-05,
+      "loss": 1.1055,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.101234155126556,
+      "learning_rate": 4.9998619317027815e-05,
+      "loss": 0.9346,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6501921982001537,
+      "learning_rate": 4.9998575138375305e-05,
+      "loss": 1.9141,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.600452999852787,
+      "learning_rate": 4.9998530264022624e-05,
+      "loss": 1.043,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.078306550999076,
+      "learning_rate": 4.999848469397102e-05,
+      "loss": 1.0745,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.407682908696776,
+      "learning_rate": 4.999843842822177e-05,
+      "loss": 0.6465,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.524749302110366,
+      "learning_rate": 4.999839146677615e-05,
+      "loss": 0.9209,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.4694033675170255,
+      "learning_rate": 4.999834380963547e-05,
+      "loss": 1.2891,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.842498419151508,
+      "learning_rate": 4.999829545680107e-05,
+      "loss": 0.8496,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8579201254124134,
+      "learning_rate": 4.9998246408274276e-05,
+      "loss": 1.4668,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5224592931908616,
+      "learning_rate": 4.9998196664056464e-05,
+      "loss": 0.5496,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.732871028912441,
+      "learning_rate": 4.999814622414901e-05,
+      "loss": 1.3359,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2115822650564696,
+      "learning_rate": 4.999809508855333e-05,
+      "loss": 0.9219,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4350131137502613,
+      "learning_rate": 4.9998043257270844e-05,
+      "loss": 1.1133,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8430125814661453,
+      "learning_rate": 4.999799073030298e-05,
+      "loss": 1.0713,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4366584441280574,
+      "learning_rate": 4.999793750765122e-05,
+      "loss": 0.8044,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9397645702946633,
+      "learning_rate": 4.999788358931703e-05,
+      "loss": 1.4355,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.554129420476975,
+      "learning_rate": 4.999782897530192e-05,
+      "loss": 1.041,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.376515729235687,
+      "learning_rate": 4.9997773665607405e-05,
+      "loss": 1.0859,
+      "step": 2172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.040154346764065,
+      "learning_rate": 4.999771766023502e-05,
+      "loss": 0.8687,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0153667754464477,
+      "learning_rate": 4.999766095918633e-05,
+      "loss": 0.6543,
+      "step": 2174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.428432618910523,
+      "learning_rate": 4.9997603562462924e-05,
+      "loss": 1.1035,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6733561197421203,
+      "learning_rate": 4.9997545470066375e-05,
+      "loss": 1.0986,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.256644718899235,
+      "learning_rate": 4.999748668199832e-05,
+      "loss": 1.3105,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4288352654758723,
+      "learning_rate": 4.9997427198260385e-05,
+      "loss": 1.4277,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2928998222694164,
+      "learning_rate": 4.999736701885422e-05,
+      "loss": 0.7363,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.73952882103331,
+      "learning_rate": 4.999730614378152e-05,
+      "loss": 1.3438,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.628398511808558,
+      "learning_rate": 4.9997244573043955e-05,
+      "loss": 1.1953,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9904455011185245,
+      "learning_rate": 4.9997182306643254e-05,
+      "loss": 0.7344,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.610773225886464,
+      "learning_rate": 4.999711934458114e-05,
+      "loss": 1.2227,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6238592260189195,
+      "learning_rate": 4.999705568685938e-05,
+      "loss": 0.8926,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1878968534068703,
+      "learning_rate": 4.9996991333479736e-05,
+      "loss": 0.5942,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1619240010231513,
+      "learning_rate": 4.9996926284444004e-05,
+      "loss": 1.4248,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.218182263719619,
+      "learning_rate": 4.999686053975398e-05,
+      "loss": 1.4668,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.349354058629276,
+      "learning_rate": 4.9996794099411506e-05,
+      "loss": 1.1562,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.638958021679555,
+      "learning_rate": 4.9996726963418425e-05,
+      "loss": 1.1289,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8362776286972227,
+      "learning_rate": 4.999665913177661e-05,
+      "loss": 1.2773,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.607363918404029,
+      "learning_rate": 4.999659060448796e-05,
+      "loss": 1.0186,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.487196828647961,
+      "learning_rate": 4.999652138155436e-05,
+      "loss": 0.666,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7081825703449294,
+      "learning_rate": 4.999645146297775e-05,
+      "loss": 1.5068,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3989690956804997,
+      "learning_rate": 4.999638084876006e-05,
+      "loss": 1.623,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1894528259912596,
+      "learning_rate": 4.999630953890327e-05,
+      "loss": 1.1387,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.082530247610696,
+      "learning_rate": 4.999623753340937e-05,
+      "loss": 0.8213,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.552031969153133,
+      "learning_rate": 4.999616483228036e-05,
+      "loss": 1.5918,
+      "step": 2197
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.790764980927626,
+      "learning_rate": 4.999609143551824e-05,
+      "loss": 0.8911,
+      "step": 2198
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.906823983718057,
+      "learning_rate": 4.9996017343125085e-05,
+      "loss": 1.5137,
+      "step": 2199
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.459571642640114,
+      "learning_rate": 4.9995942555102935e-05,
+      "loss": 1.3418,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.332126701034282,
+      "learning_rate": 4.999586707145389e-05,
+      "loss": 1.1768,
+      "step": 2201
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.005289924731206,
+      "learning_rate": 4.999579089218003e-05,
+      "loss": 2.2441,
+      "step": 2202
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.007311925569722,
+      "learning_rate": 4.999571401728349e-05,
+      "loss": 1.4375,
+      "step": 2203
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.054490228618113,
+      "learning_rate": 4.999563644676641e-05,
+      "loss": 0.7412,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.214064363396549,
+      "learning_rate": 4.999555818063093e-05,
+      "loss": 0.7178,
+      "step": 2205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.887805821204752,
+      "learning_rate": 4.999547921887925e-05,
+      "loss": 0.9082,
+      "step": 2206
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7858694054309527,
+      "learning_rate": 4.9995399561513567e-05,
+      "loss": 1.2402,
+      "step": 2207
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2404402993234975,
+      "learning_rate": 4.999531920853608e-05,
+      "loss": 0.8809,
+      "step": 2208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2058204288750147,
+      "learning_rate": 4.999523815994903e-05,
+      "loss": 1.3418,
+      "step": 2209
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.999465417915987,
+      "learning_rate": 4.999515641575468e-05,
+      "loss": 1.0283,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.743055678956449,
+      "learning_rate": 4.999507397595532e-05,
+      "loss": 0.6235,
+      "step": 2211
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.641708540221052,
+      "learning_rate": 4.999499084055321e-05,
+      "loss": 1.2363,
+      "step": 2212
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.094656562969398,
+      "learning_rate": 4.999490700955069e-05,
+      "loss": 1.7129,
+      "step": 2213
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.171475800351647,
+      "learning_rate": 4.999482248295006e-05,
+      "loss": 0.876,
+      "step": 2214
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.595944875455587,
+      "learning_rate": 4.999473726075372e-05,
+      "loss": 1.2578,
+      "step": 2215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7773293281534843,
+      "learning_rate": 4.999465134296401e-05,
+      "loss": 0.9121,
+      "step": 2216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.622957324964252,
+      "learning_rate": 4.9994564729583324e-05,
+      "loss": 0.7686,
+      "step": 2217
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.635630993406534,
+      "learning_rate": 4.9994477420614086e-05,
+      "loss": 0.7153,
+      "step": 2218
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3180512750583007,
+      "learning_rate": 4.9994389416058715e-05,
+      "loss": 1.7324,
+      "step": 2219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.141234155147326,
+      "learning_rate": 4.999430071591966e-05,
+      "loss": 0.6201,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8310083978523914,
+      "learning_rate": 4.9994211320199386e-05,
+      "loss": 1.1709,
+      "step": 2221
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9598347968317644,
+      "learning_rate": 4.9994121228900395e-05,
+      "loss": 1.7178,
+      "step": 2222
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9237291029798378,
+      "learning_rate": 4.999403044202517e-05,
+      "loss": 0.75,
+      "step": 2223
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.966827788502401,
+      "learning_rate": 4.999393895957627e-05,
+      "loss": 1.4023,
+      "step": 2224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.302777110161129,
+      "learning_rate": 4.999384678155621e-05,
+      "loss": 0.6045,
+      "step": 2225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.916860411706319,
+      "learning_rate": 4.9993753907967576e-05,
+      "loss": 0.6184,
+      "step": 2226
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.275787020168049,
+      "learning_rate": 4.9993660338812955e-05,
+      "loss": 0.8945,
+      "step": 2227
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.518731537629879,
+      "learning_rate": 4.999356607409493e-05,
+      "loss": 1.2227,
+      "step": 2228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2423476318172666,
+      "learning_rate": 4.999347111381614e-05,
+      "loss": 1.4199,
+      "step": 2229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4554453515009427,
+      "learning_rate": 4.999337545797922e-05,
+      "loss": 1.7246,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.856315768964225,
+      "learning_rate": 4.9993279106586835e-05,
+      "loss": 0.2969,
+      "step": 2231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.806402474198874,
+      "learning_rate": 4.9993182059641675e-05,
+      "loss": 1.4238,
+      "step": 2232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3323080898545494,
+      "learning_rate": 4.9993084317146425e-05,
+      "loss": 1.2285,
+      "step": 2233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.03228443231815,
+      "learning_rate": 4.9992985879103816e-05,
+      "loss": 1.2451,
+      "step": 2234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.663870904971246,
+      "learning_rate": 4.9992886745516586e-05,
+      "loss": 1.2637,
+      "step": 2235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.131640548943729,
+      "learning_rate": 4.999278691638749e-05,
+      "loss": 1.0215,
+      "step": 2236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7315297173397557,
+      "learning_rate": 4.999268639171931e-05,
+      "loss": 1.0322,
+      "step": 2237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5750952286379536,
+      "learning_rate": 4.9992585171514846e-05,
+      "loss": 0.6792,
+      "step": 2238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2979601479847074,
+      "learning_rate": 4.9992483255776906e-05,
+      "loss": 1.3184,
+      "step": 2239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4938376772197994,
+      "learning_rate": 4.999238064450833e-05,
+      "loss": 1.7207,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.43777168959824,
+      "learning_rate": 4.999227733771198e-05,
+      "loss": 1.1025,
+      "step": 2241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6353788913436986,
+      "learning_rate": 4.9992173335390724e-05,
+      "loss": 1.791,
+      "step": 2242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.280073620293595,
+      "learning_rate": 4.999206863754746e-05,
+      "loss": 1.333,
+      "step": 2243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.526457216880109,
+      "learning_rate": 4.99919632441851e-05,
+      "loss": 0.5601,
+      "step": 2244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.01713658531371,
+      "learning_rate": 4.9991857155306575e-05,
+      "loss": 1.021,
+      "step": 2245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6530074388630354,
+      "learning_rate": 4.999175037091484e-05,
+      "loss": 1.2324,
+      "step": 2246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1266329605795726,
+      "learning_rate": 4.999164289101287e-05,
+      "loss": 1.0107,
+      "step": 2247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.120974570625639,
+      "learning_rate": 4.9991534715603646e-05,
+      "loss": 0.9062,
+      "step": 2248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8506189831523487,
+      "learning_rate": 4.999142584469019e-05,
+      "loss": 0.6724,
+      "step": 2249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9644143015930244,
+      "learning_rate": 4.9991316278275525e-05,
+      "loss": 1.498,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.578149997265373,
+      "learning_rate": 4.9991206016362704e-05,
+      "loss": 0.9385,
+      "step": 2251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.594411789330753,
+      "learning_rate": 4.9991095058954783e-05,
+      "loss": 1.0225,
+      "step": 2252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.438748846713702,
+      "learning_rate": 4.999098340605487e-05,
+      "loss": 0.9734,
+      "step": 2253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.977409822906487,
+      "learning_rate": 4.999087105766607e-05,
+      "loss": 0.8914,
+      "step": 2254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.896255518248163,
+      "learning_rate": 4.9990758013791486e-05,
+      "loss": 0.9033,
+      "step": 2255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.736060519179734,
+      "learning_rate": 4.9990644274434286e-05,
+      "loss": 1.1348,
+      "step": 2256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3438965952459214,
+      "learning_rate": 4.999052983959763e-05,
+      "loss": 1.4102,
+      "step": 2257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.313726666877437,
+      "learning_rate": 4.99904147092847e-05,
+      "loss": 0.96,
+      "step": 2258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8442300074691955,
+      "learning_rate": 4.9990298883498714e-05,
+      "loss": 0.8,
+      "step": 2259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3718234053828207,
+      "learning_rate": 4.9990182362242875e-05,
+      "loss": 0.7085,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.987742017575551,
+      "learning_rate": 4.9990065145520435e-05,
+      "loss": 0.7939,
+      "step": 2261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.397305231204105,
+      "learning_rate": 4.9989947233334655e-05,
+      "loss": 1.1367,
+      "step": 2262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.983107692205546,
+      "learning_rate": 4.9989828625688815e-05,
+      "loss": 0.8086,
+      "step": 2263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.777541832304067,
+      "learning_rate": 4.998970932258622e-05,
+      "loss": 2.0449,
+      "step": 2264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1837198565651197,
+      "learning_rate": 4.998958932403019e-05,
+      "loss": 1.1133,
+      "step": 2265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.055213652038068,
+      "learning_rate": 4.998946863002406e-05,
+      "loss": 1.0166,
+      "step": 2266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6177034239616055,
+      "learning_rate": 4.9989347240571195e-05,
+      "loss": 1.1465,
+      "step": 2267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.893864646271899,
+      "learning_rate": 4.998922515567496e-05,
+      "loss": 1.3906,
+      "step": 2268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.8903007034178,
+      "learning_rate": 4.998910237533877e-05,
+      "loss": 1.085,
+      "step": 2269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.538579742951189,
+      "learning_rate": 4.998897889956603e-05,
+      "loss": 0.7891,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5836297654743734,
+      "learning_rate": 4.998885472836019e-05,
+      "loss": 1.0453,
+      "step": 2271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 27.71862999375093,
+      "learning_rate": 4.9988729861724684e-05,
+      "loss": 1.2534,
+      "step": 2272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.667907061646919,
+      "learning_rate": 4.9988604299663e-05,
+      "loss": 1.5195,
+      "step": 2273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7790868683834775,
+      "learning_rate": 4.998847804217863e-05,
+      "loss": 1.3477,
+      "step": 2274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.533411115991519,
+      "learning_rate": 4.99883510892751e-05,
+      "loss": 0.7773,
+      "step": 2275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.742432964034978,
+      "learning_rate": 4.998822344095592e-05,
+      "loss": 1.7227,
+      "step": 2276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.931074378765401,
+      "learning_rate": 4.998809509722466e-05,
+      "loss": 1.1895,
+      "step": 2277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.88497940199696,
+      "learning_rate": 4.998796605808489e-05,
+      "loss": 0.771,
+      "step": 2278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6627966594638988,
+      "learning_rate": 4.998783632354019e-05,
+      "loss": 1.2773,
+      "step": 2279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.086358227599289,
+      "learning_rate": 4.998770589359418e-05,
+      "loss": 1.2295,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5496731782294,
+      "learning_rate": 4.998757476825049e-05,
+      "loss": 1.1475,
+      "step": 2281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.592565059011071,
+      "learning_rate": 4.998744294751276e-05,
+      "loss": 0.6929,
+      "step": 2282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.960572786548942,
+      "learning_rate": 4.9987310431384673e-05,
+      "loss": 1.3242,
+      "step": 2283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3756237214889637,
+      "learning_rate": 4.9987177219869904e-05,
+      "loss": 1.2163,
+      "step": 2284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6901223451948613,
+      "learning_rate": 4.998704331297217e-05,
+      "loss": 1.1406,
+      "step": 2285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.781735951387979,
+      "learning_rate": 4.9986908710695184e-05,
+      "loss": 0.9922,
+      "step": 2286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.706958547891729,
+      "learning_rate": 4.998677341304271e-05,
+      "loss": 1.4434,
+      "step": 2287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.753972219123427,
+      "learning_rate": 4.99866374200185e-05,
+      "loss": 1.29,
+      "step": 2288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.379202331062279,
+      "learning_rate": 4.998650073162634e-05,
+      "loss": 2.0547,
+      "step": 2289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.429861557180935,
+      "learning_rate": 4.998636334787004e-05,
+      "loss": 0.8896,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.93605354019905,
+      "learning_rate": 4.998622526875342e-05,
+      "loss": 1.7012,
+      "step": 2291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5176326725083706,
+      "learning_rate": 4.9986086494280324e-05,
+      "loss": 0.9092,
+      "step": 2292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7126478796522657,
+      "learning_rate": 4.998594702445461e-05,
+      "loss": 0.7847,
+      "step": 2293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8817320151183576,
+      "learning_rate": 4.998580685928016e-05,
+      "loss": 1.3926,
+      "step": 2294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.912981870184686,
+      "learning_rate": 4.9985665998760884e-05,
+      "loss": 1.0156,
+      "step": 2295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.059428017208522,
+      "learning_rate": 4.998552444290069e-05,
+      "loss": 1.5605,
+      "step": 2296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.518614912269813,
+      "learning_rate": 4.998538219170352e-05,
+      "loss": 1.1338,
+      "step": 2297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3232350038639464,
+      "learning_rate": 4.9985239245173345e-05,
+      "loss": 1.2188,
+      "step": 2298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.631490057465607,
+      "learning_rate": 4.998509560331413e-05,
+      "loss": 1.7715,
+      "step": 2299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1449511174320723,
+      "learning_rate": 4.998495126612987e-05,
+      "loss": 1.3149,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.308822169262515,
+      "learning_rate": 4.998480623362458e-05,
+      "loss": 1.8516,
+      "step": 2301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2702125143991845,
+      "learning_rate": 4.9984660505802316e-05,
+      "loss": 1.2617,
+      "step": 2302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.530786071459982,
+      "learning_rate": 4.998451408266712e-05,
+      "loss": 1.0923,
+      "step": 2303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9084994067540195,
+      "learning_rate": 4.9984366964223075e-05,
+      "loss": 0.4744,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.834703229378419,
+      "learning_rate": 4.998421915047425e-05,
+      "loss": 0.9277,
+      "step": 2305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1576260461469605,
+      "learning_rate": 4.99840706414248e-05,
+      "loss": 1.5215,
+      "step": 2306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.367117259273177,
+      "learning_rate": 4.998392143707882e-05,
+      "loss": 1.3105,
+      "step": 2307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3104397643310355,
+      "learning_rate": 4.998377153744048e-05,
+      "loss": 0.6506,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8430273793740723,
+      "learning_rate": 4.9983620942513945e-05,
+      "loss": 1.1924,
+      "step": 2309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8785295796454053,
+      "learning_rate": 4.9983469652303415e-05,
+      "loss": 1.1934,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8844380306501123,
+      "learning_rate": 4.99833176668131e-05,
+      "loss": 0.629,
+      "step": 2311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.985545745583247,
+      "learning_rate": 4.9983164986047216e-05,
+      "loss": 1.3105,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.127631420156492,
+      "learning_rate": 4.998301161001003e-05,
+      "loss": 1.5918,
+      "step": 2313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2512993782632087,
+      "learning_rate": 4.998285753870579e-05,
+      "loss": 1.3184,
+      "step": 2314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.594783121829093,
+      "learning_rate": 4.9982702772138806e-05,
+      "loss": 1.0312,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.552768371402354,
+      "learning_rate": 4.998254731031337e-05,
+      "loss": 1.6191,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8075642138130212,
+      "learning_rate": 4.998239115323381e-05,
+      "loss": 1.2363,
+      "step": 2317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.471857689265814,
+      "learning_rate": 4.998223430090449e-05,
+      "loss": 1.4414,
+      "step": 2318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.553923397192262,
+      "learning_rate": 4.9982076753329736e-05,
+      "loss": 0.7715,
+      "step": 2319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8211424380491694,
+      "learning_rate": 4.9981918510513966e-05,
+      "loss": 1.1367,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3626943316587306,
+      "learning_rate": 4.998175957246158e-05,
+      "loss": 1.4297,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.129603966666633,
+      "learning_rate": 4.998159993917699e-05,
+      "loss": 1.1875,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.670564116224512,
+      "learning_rate": 4.9981439610664645e-05,
+      "loss": 1.3223,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3817421715361586,
+      "learning_rate": 4.9981278586929006e-05,
+      "loss": 1.6582,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.092467687710921,
+      "learning_rate": 4.9981116867974555e-05,
+      "loss": 1.3398,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.449452560921637,
+      "learning_rate": 4.9980954453805785e-05,
+      "loss": 0.9082,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.41539191115888,
+      "learning_rate": 4.998079134442722e-05,
+      "loss": 2.2227,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.415514265633815,
+      "learning_rate": 4.998062753984342e-05,
+      "loss": 1.0366,
+      "step": 2328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.123690368052201,
+      "learning_rate": 4.9980463040058904e-05,
+      "loss": 1.4141,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7310165270567532,
+      "learning_rate": 4.9980297845078285e-05,
+      "loss": 1.1582,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.639784647299924,
+      "learning_rate": 4.998013195490614e-05,
+      "loss": 1.3867,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.122358432226617,
+      "learning_rate": 4.9979965369547095e-05,
+      "loss": 1.0996,
+      "step": 2332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0694269553632436,
+      "learning_rate": 4.997979808900578e-05,
+      "loss": 0.9414,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.321289636779574,
+      "learning_rate": 4.9979630113286856e-05,
+      "loss": 1.0977,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.874988555891304,
+      "learning_rate": 4.997946144239499e-05,
+      "loss": 0.4756,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0922447164599713,
+      "learning_rate": 4.997929207633489e-05,
+      "loss": 1.0532,
+      "step": 2336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.81611893563405,
+      "learning_rate": 4.9979122015111255e-05,
+      "loss": 1.4062,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8726455335544263,
+      "learning_rate": 4.9978951258728826e-05,
+      "loss": 1.5137,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.5102057702054,
+      "learning_rate": 4.997877980719235e-05,
+      "loss": 0.9954,
+      "step": 2339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9209020290014696,
+      "learning_rate": 4.9978607660506604e-05,
+      "loss": 0.8774,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.479672059319385,
+      "learning_rate": 4.997843481867637e-05,
+      "loss": 0.7939,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.5270809690736655,
+      "learning_rate": 4.997826128170647e-05,
+      "loss": 1.1367,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.60045204503391,
+      "learning_rate": 4.997808704960173e-05,
+      "loss": 0.2363,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4404723754625324,
+      "learning_rate": 4.997791212236698e-05,
+      "loss": 1.3125,
+      "step": 2344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.600062593780569,
+      "learning_rate": 4.997773650000712e-05,
+      "loss": 1.2549,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.9347060311346,
+      "learning_rate": 4.997756018252702e-05,
+      "loss": 2.3125,
+      "step": 2346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.065590836298632,
+      "learning_rate": 4.997738316993158e-05,
+      "loss": 1.2227,
+      "step": 2347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7165620923120257,
+      "learning_rate": 4.997720546222574e-05,
+      "loss": 1.2441,
+      "step": 2348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4696864333276105,
+      "learning_rate": 4.997702705941444e-05,
+      "loss": 1.0166,
+      "step": 2349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.5246996746775086,
+      "learning_rate": 4.997684796150264e-05,
+      "loss": 1.3691,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8600386386208663,
+      "learning_rate": 4.9976668168495336e-05,
+      "loss": 0.749,
+      "step": 2351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.359249191364552,
+      "learning_rate": 4.997648768039752e-05,
+      "loss": 1.25,
+      "step": 2352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8454417405775088,
+      "learning_rate": 4.997630649721422e-05,
+      "loss": 0.8584,
+      "step": 2353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9061077452268824,
+      "learning_rate": 4.9976124618950485e-05,
+      "loss": 0.8208,
+      "step": 2354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.273806931934767,
+      "learning_rate": 4.997594204561137e-05,
+      "loss": 1.1113,
+      "step": 2355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0383527461394513,
+      "learning_rate": 4.997575877720195e-05,
+      "loss": 0.5015,
+      "step": 2356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.78764603593767,
+      "learning_rate": 4.9975574813727326e-05,
+      "loss": 1.2266,
+      "step": 2357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4580814318720425,
+      "learning_rate": 4.997539015519263e-05,
+      "loss": 1.2275,
+      "step": 2358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.916095878611606,
+      "learning_rate": 4.997520480160299e-05,
+      "loss": 1.0479,
+      "step": 2359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8715307641504895,
+      "learning_rate": 4.997501875296357e-05,
+      "loss": 1.498,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9941874461236413,
+      "learning_rate": 4.9974832009279546e-05,
+      "loss": 0.9458,
+      "step": 2361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.067063805204619,
+      "learning_rate": 4.9974644570556116e-05,
+      "loss": 0.7969,
+      "step": 2362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.99423851881868,
+      "learning_rate": 4.9974456436798486e-05,
+      "loss": 0.6689,
+      "step": 2363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0277784759775757,
+      "learning_rate": 4.997426760801191e-05,
+      "loss": 0.7297,
+      "step": 2364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.648012798234774,
+      "learning_rate": 4.9974078084201634e-05,
+      "loss": 1.2637,
+      "step": 2365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.423208841521945,
+      "learning_rate": 4.997388786537293e-05,
+      "loss": 0.6353,
+      "step": 2366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.548777423265808,
+      "learning_rate": 4.9973696951531085e-05,
+      "loss": 1.1523,
+      "step": 2367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.338791344673239,
+      "learning_rate": 4.997350534268143e-05,
+      "loss": 1.3633,
+      "step": 2368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.468964965732434,
+      "learning_rate": 4.9973313038829285e-05,
+      "loss": 1.1035,
+      "step": 2369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0154514681613933,
+      "learning_rate": 4.997312003998e-05,
+      "loss": 0.5503,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.699147379438132,
+      "learning_rate": 4.997292634613896e-05,
+      "loss": 1.2036,
+      "step": 2371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.393090909792241,
+      "learning_rate": 4.997273195731154e-05,
+      "loss": 1.2871,
+      "step": 2372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.435021151349229,
+      "learning_rate": 4.997253687350316e-05,
+      "loss": 0.8301,
+      "step": 2373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8690291059100606,
+      "learning_rate": 4.9972341094719244e-05,
+      "loss": 1.3223,
+      "step": 2374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.168274331951865,
+      "learning_rate": 4.9972144620965237e-05,
+      "loss": 1.6621,
+      "step": 2375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0121775474348302,
+      "learning_rate": 4.9971947452246615e-05,
+      "loss": 1.1992,
+      "step": 2376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.831058421486837,
+      "learning_rate": 4.997174958856886e-05,
+      "loss": 0.7778,
+      "step": 2377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.5579445004707,
+      "learning_rate": 4.9971551029937485e-05,
+      "loss": 1.3906,
+      "step": 2378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5686645790617666,
+      "learning_rate": 4.997135177635801e-05,
+      "loss": 1.3105,
+      "step": 2379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.421350926947989,
+      "learning_rate": 4.9971151827835975e-05,
+      "loss": 0.8027,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.858700681182826,
+      "learning_rate": 4.9970951184376954e-05,
+      "loss": 1.3545,
+      "step": 2381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.813187190659534,
+      "learning_rate": 4.997074984598653e-05,
+      "loss": 1.7148,
+      "step": 2382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4941697802292135,
+      "learning_rate": 4.99705478126703e-05,
+      "loss": 0.507,
+      "step": 2383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2293395129469387,
+      "learning_rate": 4.997034508443389e-05,
+      "loss": 0.9688,
+      "step": 2384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8429849621495094,
+      "learning_rate": 4.997014166128294e-05,
+      "loss": 0.7648,
+      "step": 2385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8673127274360053,
+      "learning_rate": 4.996993754322312e-05,
+      "loss": 0.9092,
+      "step": 2386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.3299859218832575,
+      "learning_rate": 4.9969732730260095e-05,
+      "loss": 0.8652,
+      "step": 2387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.808423582361699,
+      "learning_rate": 4.9969527222399577e-05,
+      "loss": 0.856,
+      "step": 2388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.478264192814638,
+      "learning_rate": 4.996932101964728e-05,
+      "loss": 1.6367,
+      "step": 2389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.72978546163391,
+      "learning_rate": 4.9969114122008944e-05,
+      "loss": 1.2441,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0147604221790507,
+      "learning_rate": 4.996890652949033e-05,
+      "loss": 1.207,
+      "step": 2391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.734843722359088,
+      "learning_rate": 4.996869824209721e-05,
+      "loss": 1.0376,
+      "step": 2392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.894616350756631,
+      "learning_rate": 4.9968489259835384e-05,
+      "loss": 1.168,
+      "step": 2393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0368151359679345,
+      "learning_rate": 4.996827958271066e-05,
+      "loss": 0.6284,
+      "step": 2394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8082336279793325,
+      "learning_rate": 4.9968069210728884e-05,
+      "loss": 1.2109,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.02857809227825,
+      "learning_rate": 4.9967858143895914e-05,
+      "loss": 0.752,
+      "step": 2396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.308810099680244,
+      "learning_rate": 4.9967646382217605e-05,
+      "loss": 0.7114,
+      "step": 2397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.404144455814294,
+      "learning_rate": 4.996743392569987e-05,
+      "loss": 0.9114,
+      "step": 2398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1014308959141945,
+      "learning_rate": 4.996722077434861e-05,
+      "loss": 1.166,
+      "step": 2399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0926886385201375,
+      "learning_rate": 4.996700692816976e-05,
+      "loss": 0.9766,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2967302191162497,
+      "learning_rate": 4.996679238716927e-05,
+      "loss": 1.4141,
+      "step": 2401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.35008066323004,
+      "learning_rate": 4.996657715135311e-05,
+      "loss": 0.6484,
+      "step": 2402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.629185382597053,
+      "learning_rate": 4.9966361220727275e-05,
+      "loss": 1.1787,
+      "step": 2403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.239034531272919,
+      "learning_rate": 4.996614459529777e-05,
+      "loss": 0.8252,
+      "step": 2404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8675080347462816,
+      "learning_rate": 4.9965927275070625e-05,
+      "loss": 1.0098,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.046683382572829,
+      "learning_rate": 4.996570926005189e-05,
+      "loss": 1.0376,
+      "step": 2406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9694269312013137,
+      "learning_rate": 4.9965490550247626e-05,
+      "loss": 1.0273,
+      "step": 2407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.306279879661509,
+      "learning_rate": 4.9965271145663925e-05,
+      "loss": 1.2812,
+      "step": 2408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5888898424720734,
+      "learning_rate": 4.9965051046306884e-05,
+      "loss": 1.0461,
+      "step": 2409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.260663238890387,
+      "learning_rate": 4.996483025218265e-05,
+      "loss": 1.0298,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9785425203703735,
+      "learning_rate": 4.996460876329735e-05,
+      "loss": 1.9609,
+      "step": 2411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7245638717078418,
+      "learning_rate": 4.996438657965714e-05,
+      "loss": 2.043,
+      "step": 2412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.138718361545294,
+      "learning_rate": 4.9964163701268224e-05,
+      "loss": 1.457,
+      "step": 2413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0094306853500976,
+      "learning_rate": 4.9963940128136793e-05,
+      "loss": 1.3125,
+      "step": 2414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.11898205244174,
+      "learning_rate": 4.9963715860269076e-05,
+      "loss": 2.1113,
+      "step": 2415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.234753581222496,
+      "learning_rate": 4.99634908976713e-05,
+      "loss": 1.0356,
+      "step": 2416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5797979302340415,
+      "learning_rate": 4.996326524034975e-05,
+      "loss": 0.7476,
+      "step": 2417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0097932656095505,
+      "learning_rate": 4.996303888831067e-05,
+      "loss": 1.2891,
+      "step": 2418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.493609749045726,
+      "learning_rate": 4.996281184156039e-05,
+      "loss": 1.0684,
+      "step": 2419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.581523435579,
+      "learning_rate": 4.996258410010522e-05,
+      "loss": 1.3945,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.412907758248806,
+      "learning_rate": 4.99623556639515e-05,
+      "loss": 1.2598,
+      "step": 2421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.508262807756527,
+      "learning_rate": 4.9962126533105574e-05,
+      "loss": 1.0898,
+      "step": 2422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2003653266328738,
+      "learning_rate": 4.9961896707573834e-05,
+      "loss": 1.2246,
+      "step": 2423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.545179399414001,
+      "learning_rate": 4.996166618736268e-05,
+      "loss": 1.0586,
+      "step": 2424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1950599823839503,
+      "learning_rate": 4.9961434972478505e-05,
+      "loss": 1.4141,
+      "step": 2425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4405569811927124,
+      "learning_rate": 4.996120306292776e-05,
+      "loss": 1.0367,
+      "step": 2426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.369436775434614,
+      "learning_rate": 4.9960970458716896e-05,
+      "loss": 1.793,
+      "step": 2427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.340847201425664,
+      "learning_rate": 4.996073715985239e-05,
+      "loss": 1.1133,
+      "step": 2428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5720326893625303,
+      "learning_rate": 4.996050316634072e-05,
+      "loss": 0.9336,
+      "step": 2429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6196061628123144,
+      "learning_rate": 4.9960268478188414e-05,
+      "loss": 0.521,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6346648713198637,
+      "learning_rate": 4.9960033095402005e-05,
+      "loss": 1.3984,
+      "step": 2431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.014760616672862,
+      "learning_rate": 4.9959797017988026e-05,
+      "loss": 1.832,
+      "step": 2432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.885194091189074,
+      "learning_rate": 4.9959560245953064e-05,
+      "loss": 1.5859,
+      "step": 2433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.744709704440817,
+      "learning_rate": 4.99593227793037e-05,
+      "loss": 1.3281,
+      "step": 2434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9571766987635284,
+      "learning_rate": 4.995908461804654e-05,
+      "loss": 0.999,
+      "step": 2435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.560797479583959,
+      "learning_rate": 4.995884576218821e-05,
+      "loss": 1.3477,
+      "step": 2436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.827767028072021,
+      "learning_rate": 4.995860621173537e-05,
+      "loss": 1.248,
+      "step": 2437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.790203874283067,
+      "learning_rate": 4.9958365966694684e-05,
+      "loss": 0.9941,
+      "step": 2438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.901144712747099,
+      "learning_rate": 4.9958125027072834e-05,
+      "loss": 2.7344,
+      "step": 2439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.707253627257894,
+      "learning_rate": 4.995788339287652e-05,
+      "loss": 0.7266,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.314563466424863,
+      "learning_rate": 4.995764106411247e-05,
+      "loss": 1.1128,
+      "step": 2441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.285897666696281,
+      "learning_rate": 4.995739804078743e-05,
+      "loss": 1.7891,
+      "step": 2442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.867496480874623,
+      "learning_rate": 4.9957154322908174e-05,
+      "loss": 1.0742,
+      "step": 2443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6717060771882304,
+      "learning_rate": 4.995690991048146e-05,
+      "loss": 0.9185,
+      "step": 2444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4048467422926656,
+      "learning_rate": 4.9956664803514106e-05,
+      "loss": 0.9697,
+      "step": 2445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.952849589957223,
+      "learning_rate": 4.995641900201293e-05,
+      "loss": 1.4004,
+      "step": 2446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.909507107672134,
+      "learning_rate": 4.995617250598478e-05,
+      "loss": 0.8525,
+      "step": 2447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.394132979237555,
+      "learning_rate": 4.99559253154365e-05,
+      "loss": 1.3359,
+      "step": 2448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.620271426821119,
+      "learning_rate": 4.995567743037498e-05,
+      "loss": 1.5156,
+      "step": 2449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4803035195943726,
+      "learning_rate": 4.9955428850807126e-05,
+      "loss": 0.7544,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.39034730417774,
+      "learning_rate": 4.9955179576739836e-05,
+      "loss": 1.4004,
+      "step": 2451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.098849562736575,
+      "learning_rate": 4.9954929608180055e-05,
+      "loss": 0.665,
+      "step": 2452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.5427267021584,
+      "learning_rate": 4.9954678945134756e-05,
+      "loss": 1.1719,
+      "step": 2453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2820318267100648,
+      "learning_rate": 4.9954427587610884e-05,
+      "loss": 0.7439,
+      "step": 2454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.333669223373901,
+      "learning_rate": 4.9954175535615465e-05,
+      "loss": 0.7944,
+      "step": 2455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0671832029727444,
+      "learning_rate": 4.9953922789155494e-05,
+      "loss": 0.96,
+      "step": 2456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0707429258489847,
+      "learning_rate": 4.995366934823801e-05,
+      "loss": 1.3496,
+      "step": 2457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.720861303656808,
+      "learning_rate": 4.995341521287007e-05,
+      "loss": 1.1826,
+      "step": 2458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.24093039757674,
+      "learning_rate": 4.9953160383058736e-05,
+      "loss": 1.3066,
+      "step": 2459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5369600025403614,
+      "learning_rate": 4.995290485881111e-05,
+      "loss": 1.5293,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5005425818070317,
+      "learning_rate": 4.9952648640134306e-05,
+      "loss": 0.8868,
+      "step": 2461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9563631207292396,
+      "learning_rate": 4.9952391727035445e-05,
+      "loss": 0.8291,
+      "step": 2462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5809493806553765,
+      "learning_rate": 4.995213411952168e-05,
+      "loss": 1.0479,
+      "step": 2463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2350116169604126,
+      "learning_rate": 4.995187581760018e-05,
+      "loss": 1.4629,
+      "step": 2464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.65030314966332,
+      "learning_rate": 4.995161682127813e-05,
+      "loss": 0.8701,
+      "step": 2465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1137883289111326,
+      "learning_rate": 4.995135713056275e-05,
+      "loss": 0.8359,
+      "step": 2466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7272898406156507,
+      "learning_rate": 4.9951096745461255e-05,
+      "loss": 1.0391,
+      "step": 2467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.567146965627409,
+      "learning_rate": 4.99508356659809e-05,
+      "loss": 1.6699,
+      "step": 2468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4304121598829154,
+      "learning_rate": 4.9950573892128936e-05,
+      "loss": 1.3867,
+      "step": 2469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9734366917571484,
+      "learning_rate": 4.995031142391267e-05,
+      "loss": 0.521,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.357972601991794,
+      "learning_rate": 4.995004826133938e-05,
+      "loss": 1.3438,
+      "step": 2471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.949021574724133,
+      "learning_rate": 4.994978440441641e-05,
+      "loss": 1.666,
+      "step": 2472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.2614414337416475,
+      "learning_rate": 4.99495198531511e-05,
+      "loss": 1.0264,
+      "step": 2473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2071384100091005,
+      "learning_rate": 4.994925460755081e-05,
+      "loss": 1.3105,
+      "step": 2474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.785852032354392,
+      "learning_rate": 4.994898866762291e-05,
+      "loss": 1.0254,
+      "step": 2475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.445863433177505,
+      "learning_rate": 4.994872203337483e-05,
+      "loss": 1.3262,
+      "step": 2476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8729491143674368,
+      "learning_rate": 4.9948454704813966e-05,
+      "loss": 0.5839,
+      "step": 2477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.607324592067998,
+      "learning_rate": 4.9948186681947756e-05,
+      "loss": 1.4609,
+      "step": 2478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1618999426719165,
+      "learning_rate": 4.9947917964783676e-05,
+      "loss": 1.1133,
+      "step": 2479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.199994586061447,
+      "learning_rate": 4.9947648553329186e-05,
+      "loss": 0.8145,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6474218530981646,
+      "learning_rate": 4.9947378447591795e-05,
+      "loss": 1.5605,
+      "step": 2481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.491135713106229,
+      "learning_rate": 4.994710764757903e-05,
+      "loss": 0.9121,
+      "step": 2482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.023399822732188,
+      "learning_rate": 4.994683615329841e-05,
+      "loss": 1.7949,
+      "step": 2483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3012810829628645,
+      "learning_rate": 4.994656396475749e-05,
+      "loss": 0.3792,
+      "step": 2484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.118312635263108,
+      "learning_rate": 4.994629108196385e-05,
+      "loss": 0.7178,
+      "step": 2485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1663787275878725,
+      "learning_rate": 4.994601750492509e-05,
+      "loss": 1.2363,
+      "step": 2486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.019130729400121,
+      "learning_rate": 4.994574323364881e-05,
+      "loss": 0.8828,
+      "step": 2487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.280996857275966,
+      "learning_rate": 4.9945468268142656e-05,
+      "loss": 1.1484,
+      "step": 2488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7999685217586565,
+      "learning_rate": 4.994519260841428e-05,
+      "loss": 1.2285,
+      "step": 2489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8180086676555687,
+      "learning_rate": 4.9944916254471344e-05,
+      "loss": 0.5266,
+      "step": 2490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.720169749011988,
+      "learning_rate": 4.9944639206321555e-05,
+      "loss": 1.2754,
+      "step": 2491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6626975402621245,
+      "learning_rate": 4.99443614639726e-05,
+      "loss": 1.4004,
+      "step": 2492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.371654267627697,
+      "learning_rate": 4.994408302743222e-05,
+      "loss": 0.9263,
+      "step": 2493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.393071851892036,
+      "learning_rate": 4.9943803896708166e-05,
+      "loss": 1.5781,
+      "step": 2494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.865439601690704,
+      "learning_rate": 4.99435240718082e-05,
+      "loss": 1.4062,
+      "step": 2495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6219374639006388,
+      "learning_rate": 4.994324355274013e-05,
+      "loss": 1.2441,
+      "step": 2496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3044055830951606,
+      "learning_rate": 4.9942962339511726e-05,
+      "loss": 0.8313,
+      "step": 2497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9423516971678727,
+      "learning_rate": 4.9942680432130843e-05,
+      "loss": 0.4604,
+      "step": 2498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1183995229893404,
+      "learning_rate": 4.994239783060532e-05,
+      "loss": 0.7964,
+      "step": 2499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.227141616981233,
+      "learning_rate": 4.9942114534943017e-05,
+      "loss": 1.1895,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.700451880002081,
+      "learning_rate": 4.994183054515183e-05,
+      "loss": 1.1074,
+      "step": 2501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.371515903081396,
+      "learning_rate": 4.9941545861239634e-05,
+      "loss": 1.293,
+      "step": 2502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3355300190686274,
+      "learning_rate": 4.994126048321438e-05,
+      "loss": 1.543,
+      "step": 2503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9429391377841831,
+      "learning_rate": 4.9940974411084e-05,
+      "loss": 0.751,
+      "step": 2504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8103708060206196,
+      "learning_rate": 4.9940687644856454e-05,
+      "loss": 0.731,
+      "step": 2505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4005497382298358,
+      "learning_rate": 4.994040018453973e-05,
+      "loss": 1.3027,
+      "step": 2506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7485641273650065,
+      "learning_rate": 4.9940112030141816e-05,
+      "loss": 1.1777,
+      "step": 2507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9498408668338403,
+      "learning_rate": 4.993982318167074e-05,
+      "loss": 1.283,
+      "step": 2508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2539607261275734,
+      "learning_rate": 4.993953363913454e-05,
+      "loss": 0.6982,
+      "step": 2509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2293740173115233,
+      "learning_rate": 4.993924340254127e-05,
+      "loss": 0.7362,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.821162299203083,
+      "learning_rate": 4.9938952471899e-05,
+      "loss": 1.2427,
+      "step": 2511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5131934679434726,
+      "learning_rate": 4.9938660847215846e-05,
+      "loss": 0.96,
+      "step": 2512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.703493477179193,
+      "learning_rate": 4.9938368528499914e-05,
+      "loss": 1.3086,
+      "step": 2513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.367958779701208,
+      "learning_rate": 4.993807551575934e-05,
+      "loss": 1.2043,
+      "step": 2514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.610112121608727,
+      "learning_rate": 4.993778180900227e-05,
+      "loss": 1.1836,
+      "step": 2515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.390388020809391,
+      "learning_rate": 4.9937487408236886e-05,
+      "loss": 0.8135,
+      "step": 2516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.818855818223395,
+      "learning_rate": 4.9937192313471374e-05,
+      "loss": 1.4834,
+      "step": 2517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2665235118912963,
+      "learning_rate": 4.993689652471396e-05,
+      "loss": 1.7266,
+      "step": 2518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5282408459602146,
+      "learning_rate": 4.993660004197287e-05,
+      "loss": 1.5684,
+      "step": 2519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.08983756684918,
+      "learning_rate": 4.993630286525634e-05,
+      "loss": 1.2275,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.186944035531362,
+      "learning_rate": 4.993600499457266e-05,
+      "loss": 1.2422,
+      "step": 2521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9171049787785073,
+      "learning_rate": 4.993570642993011e-05,
+      "loss": 0.7598,
+      "step": 2522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0203886209004045,
+      "learning_rate": 4.993540717133701e-05,
+      "loss": 1.1934,
+      "step": 2523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.691410945007924,
+      "learning_rate": 4.993510721880167e-05,
+      "loss": 1.1477,
+      "step": 2524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 133.20503022831008,
+      "learning_rate": 4.993480657233245e-05,
+      "loss": 1.3662,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.231979869757141,
+      "learning_rate": 4.9934505231937714e-05,
+      "loss": 1.291,
+      "step": 2526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.962358310319802,
+      "learning_rate": 4.993420319762585e-05,
+      "loss": 1.3125,
+      "step": 2527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1801022720685306,
+      "learning_rate": 4.993390046940526e-05,
+      "loss": 1.0273,
+      "step": 2528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.84059588561125,
+      "learning_rate": 4.993359704728437e-05,
+      "loss": 1.3008,
+      "step": 2529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.729449136283256,
+      "learning_rate": 4.993329293127163e-05,
+      "loss": 0.7852,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.930689607259131,
+      "learning_rate": 4.993298812137549e-05,
+      "loss": 0.7856,
+      "step": 2531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.980502237516466,
+      "learning_rate": 4.9932682617604446e-05,
+      "loss": 1.6094,
+      "step": 2532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5093017149374712,
+      "learning_rate": 4.9932376419966997e-05,
+      "loss": 1.2559,
+      "step": 2533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5644048904812635,
+      "learning_rate": 4.993206952847166e-05,
+      "loss": 0.7021,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3321526015000282,
+      "learning_rate": 4.9931761943126974e-05,
+      "loss": 1.5254,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.496476360481874,
+      "learning_rate": 4.993145366394151e-05,
+      "loss": 1.2939,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5616496466238976,
+      "learning_rate": 4.9931144690923834e-05,
+      "loss": 1.2227,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.761325645307301,
+      "learning_rate": 4.993083502408255e-05,
+      "loss": 0.6466,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.333086644772461,
+      "learning_rate": 4.993052466342628e-05,
+      "loss": 1.543,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.609020361754585,
+      "learning_rate": 4.993021360896366e-05,
+      "loss": 1.0415,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.714743720757241,
+      "learning_rate": 4.992990186070334e-05,
+      "loss": 1.1152,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.24420049529477,
+      "learning_rate": 4.9929589418654e-05,
+      "loss": 1.1113,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.543177272108437,
+      "learning_rate": 4.992927628282433e-05,
+      "loss": 1.4463,
+      "step": 2543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.822257382291446,
+      "learning_rate": 4.992896245322306e-05,
+      "loss": 0.6228,
+      "step": 2544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9152860575460577,
+      "learning_rate": 4.9928647929858916e-05,
+      "loss": 1.3789,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.300226478206809,
+      "learning_rate": 4.9928332712740646e-05,
+      "loss": 0.9082,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3910686604825564,
+      "learning_rate": 4.992801680187702e-05,
+      "loss": 1.248,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.32959166970226,
+      "learning_rate": 4.992770019727683e-05,
+      "loss": 0.9414,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.145891157932838,
+      "learning_rate": 4.99273828989489e-05,
+      "loss": 1.3965,
+      "step": 2549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8253943761831508,
+      "learning_rate": 4.992706490690205e-05,
+      "loss": 0.7207,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3165613863799455,
+      "learning_rate": 4.9926746221145134e-05,
+      "loss": 0.7666,
+      "step": 2551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5806416568232398,
+      "learning_rate": 4.992642684168701e-05,
+      "loss": 1.377,
+      "step": 2552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.101329746000046,
+      "learning_rate": 4.992610676853659e-05,
+      "loss": 1.1562,
+      "step": 2553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3339945296326756,
+      "learning_rate": 4.9925786001702744e-05,
+      "loss": 0.957,
+      "step": 2554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.594741477686743,
+      "learning_rate": 4.992546454119444e-05,
+      "loss": 1.5781,
+      "step": 2555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7749575465623595,
+      "learning_rate": 4.99251423870206e-05,
+      "loss": 1.3701,
+      "step": 2556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.883147760347292,
+      "learning_rate": 4.9924819539190185e-05,
+      "loss": 1.1514,
+      "step": 2557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.351751683644725,
+      "learning_rate": 4.992449599771219e-05,
+      "loss": 1.5054,
+      "step": 2558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.816983208927596,
+      "learning_rate": 4.9924171762595634e-05,
+      "loss": 1.5469,
+      "step": 2559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1332005856856338,
+      "learning_rate": 4.9923846833849513e-05,
+      "loss": 0.8672,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.534793392548273,
+      "learning_rate": 4.9923521211482885e-05,
+      "loss": 1.043,
+      "step": 2561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.128735792702861,
+      "learning_rate": 4.992319489550481e-05,
+      "loss": 1.1875,
+      "step": 2562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6659441704309446,
+      "learning_rate": 4.992286788592436e-05,
+      "loss": 0.8047,
+      "step": 2563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0359432287713792,
+      "learning_rate": 4.992254018275064e-05,
+      "loss": 0.9902,
+      "step": 2564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.606002181092504,
+      "learning_rate": 4.992221178599278e-05,
+      "loss": 1.1289,
+      "step": 2565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.791312968683706,
+      "learning_rate": 4.992188269565992e-05,
+      "loss": 1.627,
+      "step": 2566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.413513234418141,
+      "learning_rate": 4.9921552911761194e-05,
+      "loss": 1.3926,
+      "step": 2567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8537790601873856,
+      "learning_rate": 4.99212224343058e-05,
+      "loss": 1.0859,
+      "step": 2568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.598754430047854,
+      "learning_rate": 4.9920891263302936e-05,
+      "loss": 1.4004,
+      "step": 2569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.011017296122716,
+      "learning_rate": 4.992055939876181e-05,
+      "loss": 1.2129,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.567825762443761,
+      "learning_rate": 4.992022684069167e-05,
+      "loss": 1.6201,
+      "step": 2571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6828029305804506,
+      "learning_rate": 4.991989358910175e-05,
+      "loss": 0.7827,
+      "step": 2572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.890390534593888,
+      "learning_rate": 4.991955964400135e-05,
+      "loss": 0.9746,
+      "step": 2573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.581110745530349,
+      "learning_rate": 4.9919225005399737e-05,
+      "loss": 1.375,
+      "step": 2574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.86041814011808,
+      "learning_rate": 4.991888967330624e-05,
+      "loss": 0.7305,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.303520782214201,
+      "learning_rate": 4.9918553647730186e-05,
+      "loss": 0.7832,
+      "step": 2576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4108038083363024,
+      "learning_rate": 4.9918216928680936e-05,
+      "loss": 0.9863,
+      "step": 2577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.508460674262024,
+      "learning_rate": 4.9917879516167846e-05,
+      "loss": 1.3125,
+      "step": 2578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.915736738742638,
+      "learning_rate": 4.991754141020031e-05,
+      "loss": 1.2949,
+      "step": 2579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.617862299837607,
+      "learning_rate": 4.991720261078775e-05,
+      "loss": 0.8008,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.142426083468548,
+      "learning_rate": 4.991686311793958e-05,
+      "loss": 1.6152,
+      "step": 2581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.41389656484159,
+      "learning_rate": 4.9916522931665256e-05,
+      "loss": 0.9951,
+      "step": 2582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9097200412382955,
+      "learning_rate": 4.991618205197424e-05,
+      "loss": 0.7891,
+      "step": 2583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.527273712580957,
+      "learning_rate": 4.9915840478876016e-05,
+      "loss": 2.0391,
+      "step": 2584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.98743701054395,
+      "learning_rate": 4.991549821238011e-05,
+      "loss": 0.8652,
+      "step": 2585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.346674750930975,
+      "learning_rate": 4.991515525249602e-05,
+      "loss": 0.792,
+      "step": 2586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2570518611549164,
+      "learning_rate": 4.9914811599233305e-05,
+      "loss": 1.0762,
+      "step": 2587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2140349880704497,
+      "learning_rate": 4.991446725260152e-05,
+      "loss": 1.1152,
+      "step": 2588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.4581750039569625,
+      "learning_rate": 4.991412221261026e-05,
+      "loss": 1.2715,
+      "step": 2589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.202937124912638,
+      "learning_rate": 4.991377647926912e-05,
+      "loss": 1.2256,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4193087496063117,
+      "learning_rate": 4.991343005258772e-05,
+      "loss": 0.957,
+      "step": 2591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.305274992521438,
+      "learning_rate": 4.991308293257571e-05,
+      "loss": 1.1543,
+      "step": 2592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.750263204874421,
+      "learning_rate": 4.991273511924274e-05,
+      "loss": 1.0127,
+      "step": 2593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.703039796401484,
+      "learning_rate": 4.9912386612598495e-05,
+      "loss": 1.374,
+      "step": 2594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9761759189068044,
+      "learning_rate": 4.9912037412652673e-05,
+      "loss": 1.1416,
+      "step": 2595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2851678310003014,
+      "learning_rate": 4.991168751941499e-05,
+      "loss": 1.1855,
+      "step": 2596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8945929580743135,
+      "learning_rate": 4.991133693289519e-05,
+      "loss": 0.7837,
+      "step": 2597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.392838432635471,
+      "learning_rate": 4.9910985653103015e-05,
+      "loss": 1.0293,
+      "step": 2598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.040500010831833,
+      "learning_rate": 4.9910633680048255e-05,
+      "loss": 0.8428,
+      "step": 2599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.032489326628089,
+      "learning_rate": 4.991028101374071e-05,
+      "loss": 1.1953,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.826063384720089,
+      "learning_rate": 4.990992765419017e-05,
+      "loss": 1.2031,
+      "step": 2601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1083710718609656,
+      "learning_rate": 4.990957360140649e-05,
+      "loss": 0.958,
+      "step": 2602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5978659638385033,
+      "learning_rate": 4.990921885539952e-05,
+      "loss": 1.6172,
+      "step": 2603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2411822495623306,
+      "learning_rate": 4.990886341617913e-05,
+      "loss": 0.9102,
+      "step": 2604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1114121094675657,
+      "learning_rate": 4.9908507283755215e-05,
+      "loss": 0.8545,
+      "step": 2605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4297021209178755,
+      "learning_rate": 4.990815045813767e-05,
+      "loss": 0.915,
+      "step": 2606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3309839711417735,
+      "learning_rate": 4.990779293933645e-05,
+      "loss": 1.5898,
+      "step": 2607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9035009980495954,
+      "learning_rate": 4.9907434727361485e-05,
+      "loss": 1.062,
+      "step": 2608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.449793248270184,
+      "learning_rate": 4.990707582222276e-05,
+      "loss": 1.0947,
+      "step": 2609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.313923918568258,
+      "learning_rate": 4.990671622393025e-05,
+      "loss": 0.7998,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7135180686140266,
+      "learning_rate": 4.990635593249396e-05,
+      "loss": 1.2598,
+      "step": 2611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.686200287900561,
+      "learning_rate": 4.9905994947923926e-05,
+      "loss": 1.1724,
+      "step": 2612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.466946965053271,
+      "learning_rate": 4.9905633270230204e-05,
+      "loss": 1.0623,
+      "step": 2613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.494192626219846,
+      "learning_rate": 4.990527089942283e-05,
+      "loss": 0.9385,
+      "step": 2614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.363584282897266,
+      "learning_rate": 4.990490783551191e-05,
+      "loss": 0.6123,
+      "step": 2615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.649609241825839,
+      "learning_rate": 4.990454407850755e-05,
+      "loss": 1.2012,
+      "step": 2616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6094519324011713,
+      "learning_rate": 4.990417962841987e-05,
+      "loss": 1.1104,
+      "step": 2617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5200824932946646,
+      "learning_rate": 4.9903814485259e-05,
+      "loss": 0.8062,
+      "step": 2618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.484586563872128,
+      "learning_rate": 4.9903448649035125e-05,
+      "loss": 1.3008,
+      "step": 2619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.046287464514033,
+      "learning_rate": 4.99030821197584e-05,
+      "loss": 1.4141,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.355761457637018,
+      "learning_rate": 4.990271489743904e-05,
+      "loss": 0.8975,
+      "step": 2621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4465650574099107,
+      "learning_rate": 4.9902346982087266e-05,
+      "loss": 1.1348,
+      "step": 2622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7763591344649785,
+      "learning_rate": 4.990197837371331e-05,
+      "loss": 0.7213,
+      "step": 2623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.002382522564978,
+      "learning_rate": 4.9901609072327426e-05,
+      "loss": 0.8779,
+      "step": 2624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.001166752586576,
+      "learning_rate": 4.990123907793991e-05,
+      "loss": 1.3618,
+      "step": 2625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3622346436365462,
+      "learning_rate": 4.9900868390561045e-05,
+      "loss": 0.5811,
+      "step": 2626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.09080858755675,
+      "learning_rate": 4.990049701020115e-05,
+      "loss": 0.759,
+      "step": 2627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1842137768543406,
+      "learning_rate": 4.990012493687056e-05,
+      "loss": 0.8369,
+      "step": 2628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.251999460067798,
+      "learning_rate": 4.989975217057963e-05,
+      "loss": 1.4961,
+      "step": 2629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8729850114425313,
+      "learning_rate": 4.989937871133873e-05,
+      "loss": 0.9854,
+      "step": 2630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.798313700307333,
+      "learning_rate": 4.989900455915826e-05,
+      "loss": 0.9482,
+      "step": 2631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0236002145981624,
+      "learning_rate": 4.989862971404863e-05,
+      "loss": 1.2168,
+      "step": 2632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.526157707284847,
+      "learning_rate": 4.9898254176020274e-05,
+      "loss": 1.3936,
+      "step": 2633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.974879953222312,
+      "learning_rate": 4.989787794508364e-05,
+      "loss": 0.8921,
+      "step": 2634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.443534814282547,
+      "learning_rate": 4.9897501021249194e-05,
+      "loss": 1.3281,
+      "step": 2635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.869887740966875,
+      "learning_rate": 4.989712340452743e-05,
+      "loss": 1.4131,
+      "step": 2636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9244329506043676,
+      "learning_rate": 4.989674509492886e-05,
+      "loss": 1.332,
+      "step": 2637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.757818131463495,
+      "learning_rate": 4.9896366092464006e-05,
+      "loss": 1.4863,
+      "step": 2638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.309509124770984,
+      "learning_rate": 4.9895986397143425e-05,
+      "loss": 1.5195,
+      "step": 2639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.801061933986035,
+      "learning_rate": 4.9895606008977675e-05,
+      "loss": 0.6082,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.619432194490522,
+      "learning_rate": 4.989522492797735e-05,
+      "loss": 1.4414,
+      "step": 2641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8108975401789866,
+      "learning_rate": 4.989484315415304e-05,
+      "loss": 0.8652,
+      "step": 2642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2598381788660276,
+      "learning_rate": 4.9894460687515385e-05,
+      "loss": 1.2559,
+      "step": 2643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3079402405259786,
+      "learning_rate": 4.9894077528075025e-05,
+      "loss": 1.0093,
+      "step": 2644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.977759296708389,
+      "learning_rate": 4.9893693675842615e-05,
+      "loss": 0.8408,
+      "step": 2645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.441114430157247,
+      "learning_rate": 4.989330913082885e-05,
+      "loss": 1.0469,
+      "step": 2646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.502678675870513,
+      "learning_rate": 4.989292389304442e-05,
+      "loss": 0.667,
+      "step": 2647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.099312068632389,
+      "learning_rate": 4.9892537962500064e-05,
+      "loss": 0.9629,
+      "step": 2648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8135763178507345,
+      "learning_rate": 4.98921513392065e-05,
+      "loss": 0.9434,
+      "step": 2649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4330405699957605,
+      "learning_rate": 4.98917640231745e-05,
+      "loss": 1.293,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.513909222154922,
+      "learning_rate": 4.989137601441484e-05,
+      "loss": 1.1221,
+      "step": 2651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.079316531510862,
+      "learning_rate": 4.989098731293832e-05,
+      "loss": 1.0098,
+      "step": 2652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.314819817698544,
+      "learning_rate": 4.989059791875576e-05,
+      "loss": 1.3188,
+      "step": 2653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1479511612182907,
+      "learning_rate": 4.989020783187799e-05,
+      "loss": 0.8799,
+      "step": 2654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3869679722005652,
+      "learning_rate": 4.988981705231587e-05,
+      "loss": 0.7944,
+      "step": 2655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.84467627562402,
+      "learning_rate": 4.9889425580080274e-05,
+      "loss": 0.8115,
+      "step": 2656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.4615434520529185,
+      "learning_rate": 4.988903341518209e-05,
+      "loss": 1.541,
+      "step": 2657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7625977063053435,
+      "learning_rate": 4.9888640557632247e-05,
+      "loss": 1.4551,
+      "step": 2658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5858846699368865,
+      "learning_rate": 4.988824700744166e-05,
+      "loss": 0.647,
+      "step": 2659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.662233840277632,
+      "learning_rate": 4.98878527646213e-05,
+      "loss": 1.4277,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.448484271101221,
+      "learning_rate": 4.988745782918213e-05,
+      "loss": 1.2598,
+      "step": 2661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.478979690127006,
+      "learning_rate": 4.988706220113513e-05,
+      "loss": 1.5078,
+      "step": 2662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.502403253323939,
+      "learning_rate": 4.988666588049133e-05,
+      "loss": 1.1387,
+      "step": 2663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.725708223754869,
+      "learning_rate": 4.9886268867261744e-05,
+      "loss": 0.8936,
+      "step": 2664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.908152411882357,
+      "learning_rate": 4.988587116145743e-05,
+      "loss": 1.3594,
+      "step": 2665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.724074697829597,
+      "learning_rate": 4.988547276308945e-05,
+      "loss": 1.8535,
+      "step": 2666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.591302773260495,
+      "learning_rate": 4.9885073672168905e-05,
+      "loss": 0.957,
+      "step": 2667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.462296126660596,
+      "learning_rate": 4.988467388870688e-05,
+      "loss": 1.4668,
+      "step": 2668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.11360669830657,
+      "learning_rate": 4.988427341271451e-05,
+      "loss": 0.939,
+      "step": 2669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.17977027879042,
+      "learning_rate": 4.988387224420295e-05,
+      "loss": 1.0176,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.015090025258383,
+      "learning_rate": 4.988347038318334e-05,
+      "loss": 0.4253,
+      "step": 2671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.434362593142977,
+      "learning_rate": 4.988306782966689e-05,
+      "loss": 1.4268,
+      "step": 2672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.580698701349973,
+      "learning_rate": 4.9882664583664794e-05,
+      "loss": 1.2637,
+      "step": 2673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7029050608802514,
+      "learning_rate": 4.988226064518827e-05,
+      "loss": 0.9287,
+      "step": 2674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.145009498698639,
+      "learning_rate": 4.988185601424856e-05,
+      "loss": 1.1084,
+      "step": 2675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.451304236196243,
+      "learning_rate": 4.988145069085692e-05,
+      "loss": 0.709,
+      "step": 2676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.017693093704075,
+      "learning_rate": 4.9881044675024645e-05,
+      "loss": 1.3516,
+      "step": 2677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3942012648423376,
+      "learning_rate": 4.988063796676303e-05,
+      "loss": 0.698,
+      "step": 2678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.001623777779056,
+      "learning_rate": 4.988023056608339e-05,
+      "loss": 1.286,
+      "step": 2679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.929389307741222,
+      "learning_rate": 4.987982247299705e-05,
+      "loss": 1.1172,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.653267415700987,
+      "learning_rate": 4.9879413687515384e-05,
+      "loss": 1.1348,
+      "step": 2681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4325314767482014,
+      "learning_rate": 4.9879004209649774e-05,
+      "loss": 1.7559,
+      "step": 2682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.744009689519696,
+      "learning_rate": 4.9878594039411587e-05,
+      "loss": 1.6172,
+      "step": 2683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.558806505117171,
+      "learning_rate": 4.9878183176812266e-05,
+      "loss": 1.043,
+      "step": 2684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.210028795253218,
+      "learning_rate": 4.987777162186324e-05,
+      "loss": 1.02,
+      "step": 2685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.208259746814049,
+      "learning_rate": 4.987735937457595e-05,
+      "loss": 1.3701,
+      "step": 2686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.403047486482537,
+      "learning_rate": 4.987694643496188e-05,
+      "loss": 1.1816,
+      "step": 2687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.979505630807922,
+      "learning_rate": 4.987653280303251e-05,
+      "loss": 0.8555,
+      "step": 2688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5440840143330297,
+      "learning_rate": 4.9876118478799364e-05,
+      "loss": 1.2539,
+      "step": 2689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6092553025619694,
+      "learning_rate": 4.987570346227397e-05,
+      "loss": 0.874,
+      "step": 2690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.2606666448079125,
+      "learning_rate": 4.987528775346788e-05,
+      "loss": 0.8833,
+      "step": 2691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0502383717425565,
+      "learning_rate": 4.987487135239265e-05,
+      "loss": 0.9727,
+      "step": 2692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.086392196276337,
+      "learning_rate": 4.9874454259059876e-05,
+      "loss": 1.1582,
+      "step": 2693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.5283989717464666,
+      "learning_rate": 4.987403647348118e-05,
+      "loss": 1.6338,
+      "step": 2694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8606487834759347,
+      "learning_rate": 4.9873617995668164e-05,
+      "loss": 1.5254,
+      "step": 2695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.3427419325119825,
+      "learning_rate": 4.987319882563248e-05,
+      "loss": 1.3184,
+      "step": 2696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.590665472083088,
+      "learning_rate": 4.987277896338581e-05,
+      "loss": 0.9902,
+      "step": 2697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.056851730033112,
+      "learning_rate": 4.987235840893982e-05,
+      "loss": 1.3945,
+      "step": 2698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.584715923934567,
+      "learning_rate": 4.9871937162306225e-05,
+      "loss": 0.4092,
+      "step": 2699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.780743880928694,
+      "learning_rate": 4.9871515223496744e-05,
+      "loss": 0.8909,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.80686360470912,
+      "learning_rate": 4.987109259252312e-05,
+      "loss": 1.2578,
+      "step": 2701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7735422034043293,
+      "learning_rate": 4.9870669269397106e-05,
+      "loss": 1.3965,
+      "step": 2702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6961160097886228,
+      "learning_rate": 4.9870245254130496e-05,
+      "loss": 1.376,
+      "step": 2703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2553915488254326,
+      "learning_rate": 4.986982054673509e-05,
+      "loss": 0.498,
+      "step": 2704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.247006844661703,
+      "learning_rate": 4.98693951472227e-05,
+      "loss": 1.9688,
+      "step": 2705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.232231823430236,
+      "learning_rate": 4.986896905560517e-05,
+      "loss": 1.1084,
+      "step": 2706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.190527972849287,
+      "learning_rate": 4.9868542271894344e-05,
+      "loss": 0.9902,
+      "step": 2707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.115035076403102,
+      "learning_rate": 4.9868114796102126e-05,
+      "loss": 1.0938,
+      "step": 2708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.818616954429445,
+      "learning_rate": 4.986768662824039e-05,
+      "loss": 0.9136,
+      "step": 2709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0300907395879615,
+      "learning_rate": 4.986725776832106e-05,
+      "loss": 1.5566,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1050538367673486,
+      "learning_rate": 4.9866828216356065e-05,
+      "loss": 1.3295,
+      "step": 2711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.678569388951174,
+      "learning_rate": 4.986639797235737e-05,
+      "loss": 0.542,
+      "step": 2712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1028776258404385,
+      "learning_rate": 4.986596703633694e-05,
+      "loss": 0.7236,
+      "step": 2713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.312937144719431,
+      "learning_rate": 4.9865535408306765e-05,
+      "loss": 1.4121,
+      "step": 2714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1909657504671367,
+      "learning_rate": 4.986510308827888e-05,
+      "loss": 0.8555,
+      "step": 2715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.898937491421441,
+      "learning_rate": 4.9864670076265276e-05,
+      "loss": 1.2529,
+      "step": 2716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.817219776573262,
+      "learning_rate": 4.986423637227804e-05,
+      "loss": 1.5332,
+      "step": 2717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5599159104349187,
+      "learning_rate": 4.9863801976329225e-05,
+      "loss": 0.8813,
+      "step": 2718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.859898405376854,
+      "learning_rate": 4.986336688843092e-05,
+      "loss": 0.9287,
+      "step": 2719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.08373651360082,
+      "learning_rate": 4.986293110859524e-05,
+      "loss": 1.3359,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.032567931819485,
+      "learning_rate": 4.986249463683431e-05,
+      "loss": 1.1328,
+      "step": 2721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.381350441391001,
+      "learning_rate": 4.9862057473160276e-05,
+      "loss": 1.0693,
+      "step": 2722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.21549848050641,
+      "learning_rate": 4.98616196175853e-05,
+      "loss": 1.3105,
+      "step": 2723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.634267336070116,
+      "learning_rate": 4.9861181070121566e-05,
+      "loss": 1.2783,
+      "step": 2724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7781885843020597,
+      "learning_rate": 4.986074183078129e-05,
+      "loss": 0.9189,
+      "step": 2725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.760246002871557,
+      "learning_rate": 4.9860301899576695e-05,
+      "loss": 1.373,
+      "step": 2726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9319931759081403,
+      "learning_rate": 4.985986127652002e-05,
+      "loss": 0.6099,
+      "step": 2727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8165148583156876,
+      "learning_rate": 4.985941996162351e-05,
+      "loss": 1.1562,
+      "step": 2728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.35916038723241,
+      "learning_rate": 4.985897795489948e-05,
+      "loss": 1.0029,
+      "step": 2729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7984588939942,
+      "learning_rate": 4.9858535256360206e-05,
+      "loss": 0.9443,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2547716178407597,
+      "learning_rate": 4.985809186601801e-05,
+      "loss": 1.3125,
+      "step": 2731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.219554683095458,
+      "learning_rate": 4.985764778388524e-05,
+      "loss": 1.1074,
+      "step": 2732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.421521730346442,
+      "learning_rate": 4.985720300997426e-05,
+      "loss": 0.9631,
+      "step": 2733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.372901089432662,
+      "learning_rate": 4.985675754429744e-05,
+      "loss": 1.3428,
+      "step": 2734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7242017261843445,
+      "learning_rate": 4.985631138686716e-05,
+      "loss": 1.3086,
+      "step": 2735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6108163290710986,
+      "learning_rate": 4.985586453769587e-05,
+      "loss": 1.0264,
+      "step": 2736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.388600819308309,
+      "learning_rate": 4.985541699679598e-05,
+      "loss": 1.8525,
+      "step": 2737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.426907624995957,
+      "learning_rate": 4.985496876417996e-05,
+      "loss": 0.9502,
+      "step": 2738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.126260383665822,
+      "learning_rate": 4.985451983986027e-05,
+      "loss": 1.082,
+      "step": 2739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9320474626267357,
+      "learning_rate": 4.985407022384943e-05,
+      "loss": 1.125,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4195897303144625,
+      "learning_rate": 4.985361991615991e-05,
+      "loss": 1.0645,
+      "step": 2741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.04664561453581,
+      "learning_rate": 4.9853168916804276e-05,
+      "loss": 1.3779,
+      "step": 2742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.709345855387949,
+      "learning_rate": 4.985271722579507e-05,
+      "loss": 1.0352,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.300724909374934,
+      "learning_rate": 4.985226484314486e-05,
+      "loss": 1.3047,
+      "step": 2744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.888844187984713,
+      "learning_rate": 4.985181176886623e-05,
+      "loss": 0.8916,
+      "step": 2745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9716770876847953,
+      "learning_rate": 4.9851358002971805e-05,
+      "loss": 1.2402,
+      "step": 2746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.302992267554809,
+      "learning_rate": 4.98509035454742e-05,
+      "loss": 2.0547,
+      "step": 2747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.64716531688727,
+      "learning_rate": 4.985044839638606e-05,
+      "loss": 1.9512,
+      "step": 2748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.826995118090068,
+      "learning_rate": 4.984999255572007e-05,
+      "loss": 1.1377,
+      "step": 2749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.009686899875622,
+      "learning_rate": 4.98495360234889e-05,
+      "loss": 1.1875,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.131097841039605,
+      "learning_rate": 4.984907879970525e-05,
+      "loss": 1.6523,
+      "step": 2751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4761752719377434,
+      "learning_rate": 4.9848620884381866e-05,
+      "loss": 0.6401,
+      "step": 2752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1699625280985964,
+      "learning_rate": 4.984816227753147e-05,
+      "loss": 1.0264,
+      "step": 2753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7200575381105367,
+      "learning_rate": 4.9847702979166834e-05,
+      "loss": 1.0947,
+      "step": 2754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.647514700655147,
+      "learning_rate": 4.9847242989300744e-05,
+      "loss": 1.4922,
+      "step": 2755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.476963418047237,
+      "learning_rate": 4.984678230794599e-05,
+      "loss": 0.9297,
+      "step": 2756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9059686011805708,
+      "learning_rate": 4.984632093511541e-05,
+      "loss": 0.7192,
+      "step": 2757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.7296169027472965,
+      "learning_rate": 4.984585887082183e-05,
+      "loss": 0.6045,
+      "step": 2758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9505161409274923,
+      "learning_rate": 4.9845396115078104e-05,
+      "loss": 0.9082,
+      "step": 2759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.466754107850882,
+      "learning_rate": 4.984493266789713e-05,
+      "loss": 1.0039,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.191700878574152,
+      "learning_rate": 4.984446852929178e-05,
+      "loss": 1.042,
+      "step": 2761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.067335940673818,
+      "learning_rate": 4.9844003699274996e-05,
+      "loss": 1.2893,
+      "step": 2762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4062165424908586,
+      "learning_rate": 4.98435381778597e-05,
+      "loss": 1.1016,
+      "step": 2763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.566355061710069,
+      "learning_rate": 4.9843071965058855e-05,
+      "loss": 1.0576,
+      "step": 2764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.797380704129778,
+      "learning_rate": 4.984260506088542e-05,
+      "loss": 1.2656,
+      "step": 2765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.25451461541571,
+      "learning_rate": 4.984213746535241e-05,
+      "loss": 1.5684,
+      "step": 2766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.7034126104704574,
+      "learning_rate": 4.984166917847282e-05,
+      "loss": 1.3418,
+      "step": 2767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9312430224246295,
+      "learning_rate": 4.9841200200259697e-05,
+      "loss": 1.0459,
+      "step": 2768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.130403492637691,
+      "learning_rate": 4.984073053072608e-05,
+      "loss": 1.002,
+      "step": 2769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.300384124523843,
+      "learning_rate": 4.984026016988504e-05,
+      "loss": 0.2822,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6887593691641498,
+      "learning_rate": 4.983978911774969e-05,
+      "loss": 0.8369,
+      "step": 2771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.628440402125782,
+      "learning_rate": 4.983931737433311e-05,
+      "loss": 0.605,
+      "step": 2772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8768377199132156,
+      "learning_rate": 4.983884493964843e-05,
+      "loss": 0.78,
+      "step": 2773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.092898057709451,
+      "learning_rate": 4.9838371813708816e-05,
+      "loss": 0.4692,
+      "step": 2774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.953284526734532,
+      "learning_rate": 4.983789799652743e-05,
+      "loss": 1.666,
+      "step": 2775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.197943241037963,
+      "learning_rate": 4.9837423488117455e-05,
+      "loss": 1.4102,
+      "step": 2776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.086910235781994,
+      "learning_rate": 4.983694828849209e-05,
+      "loss": 1.4233,
+      "step": 2777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.373728104040782,
+      "learning_rate": 4.983647239766457e-05,
+      "loss": 0.8115,
+      "step": 2778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.186018285195635,
+      "learning_rate": 4.983599581564813e-05,
+      "loss": 1.2559,
+      "step": 2779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.44904492640778,
+      "learning_rate": 4.983551854245604e-05,
+      "loss": 2.043,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.913428462711906,
+      "learning_rate": 4.9835040578101575e-05,
+      "loss": 1.4199,
+      "step": 2781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.779997039765902,
+      "learning_rate": 4.983456192259804e-05,
+      "loss": 0.9199,
+      "step": 2782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.237697939915686,
+      "learning_rate": 4.9834082575958764e-05,
+      "loss": 0.6475,
+      "step": 2783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.604209089493905,
+      "learning_rate": 4.983360253819708e-05,
+      "loss": 1.1475,
+      "step": 2784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.2131257045275685,
+      "learning_rate": 4.983312180932634e-05,
+      "loss": 1.3828,
+      "step": 2785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1570365605525583,
+      "learning_rate": 4.983264038935993e-05,
+      "loss": 1.1914,
+      "step": 2786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.613599593429852,
+      "learning_rate": 4.9832158278311254e-05,
+      "loss": 1.7656,
+      "step": 2787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.104081237493253,
+      "learning_rate": 4.9831675476193714e-05,
+      "loss": 1.2031,
+      "step": 2788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.616637350804458,
+      "learning_rate": 4.983119198302077e-05,
+      "loss": 1.1929,
+      "step": 2789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4770293533099914,
+      "learning_rate": 4.983070779880584e-05,
+      "loss": 0.563,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.610310304656881,
+      "learning_rate": 4.983022292356244e-05,
+      "loss": 1.0107,
+      "step": 2791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2350090567727463,
+      "learning_rate": 4.9829737357304026e-05,
+      "loss": 0.8037,
+      "step": 2792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.253057043939726,
+      "learning_rate": 4.9829251100044146e-05,
+      "loss": 1.7168,
+      "step": 2793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.473607113426513,
+      "learning_rate": 4.982876415179631e-05,
+      "loss": 1.2822,
+      "step": 2794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5509624796537715,
+      "learning_rate": 4.982827651257407e-05,
+      "loss": 0.8818,
+      "step": 2795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.738501002572186,
+      "learning_rate": 4.982778818239101e-05,
+      "loss": 0.7974,
+      "step": 2796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9523225280170051,
+      "learning_rate": 4.982729916126071e-05,
+      "loss": 1.1133,
+      "step": 2797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.706146810037518,
+      "learning_rate": 4.9826809449196785e-05,
+      "loss": 1.2607,
+      "step": 2798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.7709039185923094,
+      "learning_rate": 4.982631904621286e-05,
+      "loss": 1.0654,
+      "step": 2799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.819576803767774,
+      "learning_rate": 4.9825827952322565e-05,
+      "loss": 2.2109,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.339960416964399,
+      "learning_rate": 4.9825336167539604e-05,
+      "loss": 1.4531,
+      "step": 2801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.828413300546095,
+      "learning_rate": 4.982484369187764e-05,
+      "loss": 1.0625,
+      "step": 2802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.668307945570734,
+      "learning_rate": 4.982435052535038e-05,
+      "loss": 1.5205,
+      "step": 2803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5285139956351927,
+      "learning_rate": 4.982385666797155e-05,
+      "loss": 0.8354,
+      "step": 2804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.837657996577889,
+      "learning_rate": 4.9823362119754905e-05,
+      "loss": 0.8193,
+      "step": 2805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.930443261894289,
+      "learning_rate": 4.982286688071418e-05,
+      "loss": 1.0898,
+      "step": 2806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.185625744743973,
+      "learning_rate": 4.9822370950863185e-05,
+      "loss": 0.915,
+      "step": 2807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.597270369012625,
+      "learning_rate": 4.9821874330215714e-05,
+      "loss": 0.8848,
+      "step": 2808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8351455603718974,
+      "learning_rate": 4.982137701878558e-05,
+      "loss": 0.75,
+      "step": 2809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.143525342187802,
+      "learning_rate": 4.982087901658663e-05,
+      "loss": 1.1816,
+      "step": 2810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.904587773879621,
+      "learning_rate": 4.982038032363272e-05,
+      "loss": 1.0671,
+      "step": 2811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6736247337455694,
+      "learning_rate": 4.9819880939937736e-05,
+      "loss": 1.4688,
+      "step": 2812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.62836931832584,
+      "learning_rate": 4.981938086551556e-05,
+      "loss": 1.4941,
+      "step": 2813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7966638192181974,
+      "learning_rate": 4.981888010038013e-05,
+      "loss": 1.0549,
+      "step": 2814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.845109048908865,
+      "learning_rate": 4.981837864454537e-05,
+      "loss": 1.0884,
+      "step": 2815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6226546494227647,
+      "learning_rate": 4.981787649802523e-05,
+      "loss": 0.833,
+      "step": 2816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9950367019260222,
+      "learning_rate": 4.981737366083369e-05,
+      "loss": 0.9688,
+      "step": 2817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.202155018386827,
+      "learning_rate": 4.981687013298474e-05,
+      "loss": 1.4766,
+      "step": 2818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.929279786701376,
+      "learning_rate": 4.9816365914492403e-05,
+      "loss": 1.3789,
+      "step": 2819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.851245601237582,
+      "learning_rate": 4.981586100537071e-05,
+      "loss": 0.8955,
+      "step": 2820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.263745965086795,
+      "learning_rate": 4.98153554056337e-05,
+      "loss": 1.3047,
+      "step": 2821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.274561876036333,
+      "learning_rate": 4.981484911529545e-05,
+      "loss": 0.4443,
+      "step": 2822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6301259597488595,
+      "learning_rate": 4.981434213437006e-05,
+      "loss": 1.147,
+      "step": 2823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.96596282274234,
+      "learning_rate": 4.9813834462871624e-05,
+      "loss": 1.2969,
+      "step": 2824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.83725373800154,
+      "learning_rate": 4.981332610081428e-05,
+      "loss": 1.0131,
+      "step": 2825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.389437779703819,
+      "learning_rate": 4.9812817048212165e-05,
+      "loss": 1.2773,
+      "step": 2826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.600302093990365,
+      "learning_rate": 4.9812307305079465e-05,
+      "loss": 1.1455,
+      "step": 2827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.718511022829206,
+      "learning_rate": 4.981179687143034e-05,
+      "loss": 1.3594,
+      "step": 2828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.686836959310473,
+      "learning_rate": 4.9811285747279016e-05,
+      "loss": 1.1655,
+      "step": 2829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.498837141592016,
+      "learning_rate": 4.981077393263971e-05,
+      "loss": 0.8643,
+      "step": 2830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.240323679733157,
+      "learning_rate": 4.981026142752667e-05,
+      "loss": 1.0449,
+      "step": 2831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.636069171142633,
+      "learning_rate": 4.980974823195415e-05,
+      "loss": 1.2441,
+      "step": 2832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8977568234223128,
+      "learning_rate": 4.980923434593644e-05,
+      "loss": 0.8966,
+      "step": 2833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.265976501494047,
+      "learning_rate": 4.9808719769487835e-05,
+      "loss": 1.6777,
+      "step": 2834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.616749549537146,
+      "learning_rate": 4.9808204502622655e-05,
+      "loss": 1.7188,
+      "step": 2835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8747925061524313,
+      "learning_rate": 4.9807688545355246e-05,
+      "loss": 1.1201,
+      "step": 2836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.455271421099836,
+      "learning_rate": 4.9807171897699965e-05,
+      "loss": 1.0205,
+      "step": 2837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.72745191173312,
+      "learning_rate": 4.98066545596712e-05,
+      "loss": 0.824,
+      "step": 2838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5950154354393455,
+      "learning_rate": 4.980613653128332e-05,
+      "loss": 1.4531,
+      "step": 2839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.657471412030691,
+      "learning_rate": 4.980561781255077e-05,
+      "loss": 0.748,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.641733593126896,
+      "learning_rate": 4.980509840348796e-05,
+      "loss": 1.0615,
+      "step": 2841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9713293816181645,
+      "learning_rate": 4.980457830410937e-05,
+      "loss": 1.1797,
+      "step": 2842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3296885520477275,
+      "learning_rate": 4.980405751442947e-05,
+      "loss": 1.0684,
+      "step": 2843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.665358421170718,
+      "learning_rate": 4.980353603446274e-05,
+      "loss": 1.1318,
+      "step": 2844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.093889854711104,
+      "learning_rate": 4.9803013864223704e-05,
+      "loss": 1.2593,
+      "step": 2845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.784446936297826,
+      "learning_rate": 4.980249100372688e-05,
+      "loss": 1.5742,
+      "step": 2846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.364890926086339,
+      "learning_rate": 4.980196745298683e-05,
+      "loss": 1.4932,
+      "step": 2847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1700542694548335,
+      "learning_rate": 4.980144321201813e-05,
+      "loss": 1.459,
+      "step": 2848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.122654565952138,
+      "learning_rate": 4.980091828083536e-05,
+      "loss": 1.375,
+      "step": 2849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.642389745673465,
+      "learning_rate": 4.9800392659453134e-05,
+      "loss": 1.627,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.356817307563454,
+      "learning_rate": 4.979986634788607e-05,
+      "loss": 1.4688,
+      "step": 2851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6735114724444773,
+      "learning_rate": 4.979933934614882e-05,
+      "loss": 0.9912,
+      "step": 2852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6975766609619005,
+      "learning_rate": 4.979881165425606e-05,
+      "loss": 0.9385,
+      "step": 2853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8173509597913218,
+      "learning_rate": 4.979828327222246e-05,
+      "loss": 1.0342,
+      "step": 2854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.942752410509914,
+      "learning_rate": 4.979775420006274e-05,
+      "loss": 0.8457,
+      "step": 2855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.433479024714782,
+      "learning_rate": 4.9797224437791605e-05,
+      "loss": 1.2188,
+      "step": 2856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.408078883838635,
+      "learning_rate": 4.979669398542382e-05,
+      "loss": 1.1309,
+      "step": 2857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9983643205309143,
+      "learning_rate": 4.979616284297413e-05,
+      "loss": 1.0225,
+      "step": 2858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.372044151949276,
+      "learning_rate": 4.979563101045732e-05,
+      "loss": 1.0996,
+      "step": 2859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.732404141268878,
+      "learning_rate": 4.9795098487888194e-05,
+      "loss": 1.4414,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1725571473641914,
+      "learning_rate": 4.979456527528158e-05,
+      "loss": 1.2559,
+      "step": 2861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7282307018216265,
+      "learning_rate": 4.97940313726523e-05,
+      "loss": 0.7786,
+      "step": 2862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.119327228022003,
+      "learning_rate": 4.979349678001522e-05,
+      "loss": 1.2002,
+      "step": 2863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.800353290204162,
+      "learning_rate": 4.979296149738522e-05,
+      "loss": 1.0801,
+      "step": 2864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.163170145087788,
+      "learning_rate": 4.97924255247772e-05,
+      "loss": 0.8894,
+      "step": 2865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.664958136763103,
+      "learning_rate": 4.979188886220606e-05,
+      "loss": 0.9492,
+      "step": 2866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7436861170069986,
+      "learning_rate": 4.979135150968676e-05,
+      "loss": 1.2461,
+      "step": 2867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.130527376581761,
+      "learning_rate": 4.979081346723422e-05,
+      "loss": 1.0488,
+      "step": 2868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8546029409850313,
+      "learning_rate": 4.979027473486345e-05,
+      "loss": 0.9138,
+      "step": 2869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2090091990619976,
+      "learning_rate": 4.978973531258942e-05,
+      "loss": 0.6665,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.503279918586725,
+      "learning_rate": 4.978919520042715e-05,
+      "loss": 0.5366,
+      "step": 2871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8473308780272246,
+      "learning_rate": 4.978865439839167e-05,
+      "loss": 1.0156,
+      "step": 2872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.480417452504404,
+      "learning_rate": 4.978811290649803e-05,
+      "loss": 1.5254,
+      "step": 2873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0370591334268817,
+      "learning_rate": 4.97875707247613e-05,
+      "loss": 0.9199,
+      "step": 2874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.477979399183488,
+      "learning_rate": 4.978702785319656e-05,
+      "loss": 1.9082,
+      "step": 2875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.457871435201477,
+      "learning_rate": 4.9786484291818935e-05,
+      "loss": 1.5898,
+      "step": 2876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4842399263074157,
+      "learning_rate": 4.978594004064354e-05,
+      "loss": 1.1602,
+      "step": 2877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.037717523861912,
+      "learning_rate": 4.978539509968552e-05,
+      "loss": 1.1875,
+      "step": 2878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.495011311354444,
+      "learning_rate": 4.978484946896005e-05,
+      "loss": 1.4014,
+      "step": 2879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.235776774791638,
+      "learning_rate": 4.9784303148482306e-05,
+      "loss": 0.6409,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.921143989790314,
+      "learning_rate": 4.978375613826749e-05,
+      "loss": 1.1973,
+      "step": 2881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4346501414610313,
+      "learning_rate": 4.978320843833084e-05,
+      "loss": 0.7881,
+      "step": 2882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3154293785915545,
+      "learning_rate": 4.978266004868758e-05,
+      "loss": 1.2539,
+      "step": 2883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1181348541831975,
+      "learning_rate": 4.978211096935299e-05,
+      "loss": 1.2334,
+      "step": 2884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.836221408631796,
+      "learning_rate": 4.978156120034233e-05,
+      "loss": 0.625,
+      "step": 2885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6305476105178505,
+      "learning_rate": 4.9781010741670916e-05,
+      "loss": 1.0029,
+      "step": 2886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7839642075747153,
+      "learning_rate": 4.978045959335406e-05,
+      "loss": 0.7563,
+      "step": 2887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2902204169261675,
+      "learning_rate": 4.9779907755407096e-05,
+      "loss": 0.7427,
+      "step": 2888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.648153734337112,
+      "learning_rate": 4.9779355227845395e-05,
+      "loss": 0.7622,
+      "step": 2889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.568312499854838,
+      "learning_rate": 4.977880201068432e-05,
+      "loss": 1.7793,
+      "step": 2890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.55080527138024,
+      "learning_rate": 4.9778248103939286e-05,
+      "loss": 1.3535,
+      "step": 2891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.486084355992343,
+      "learning_rate": 4.977769350762568e-05,
+      "loss": 1.6582,
+      "step": 2892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.027233805967291,
+      "learning_rate": 4.977713822175896e-05,
+      "loss": 1.1367,
+      "step": 2893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.574047030798049,
+      "learning_rate": 4.9776582246354564e-05,
+      "loss": 1.2305,
+      "step": 2894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4142102440125055,
+      "learning_rate": 4.9776025581427966e-05,
+      "loss": 0.9731,
+      "step": 2895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.441735172898434,
+      "learning_rate": 4.977546822699467e-05,
+      "loss": 1.4199,
+      "step": 2896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7736019461865635,
+      "learning_rate": 4.977491018307018e-05,
+      "loss": 1.21,
+      "step": 2897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.452939967647864,
+      "learning_rate": 4.977435144967002e-05,
+      "loss": 2.1211,
+      "step": 2898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5050126367429164,
+      "learning_rate": 4.977379202680975e-05,
+      "loss": 0.7251,
+      "step": 2899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7115911305916263,
+      "learning_rate": 4.9773231914504933e-05,
+      "loss": 1.2266,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9590842315690375,
+      "learning_rate": 4.9772671112771154e-05,
+      "loss": 0.7793,
+      "step": 2901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.594180132392197,
+      "learning_rate": 4.9772109621624024e-05,
+      "loss": 1.051,
+      "step": 2902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.434229047545579,
+      "learning_rate": 4.977154744107917e-05,
+      "loss": 0.8844,
+      "step": 2903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4756169484620787,
+      "learning_rate": 4.977098457115223e-05,
+      "loss": 1.3623,
+      "step": 2904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8633907773757452,
+      "learning_rate": 4.977042101185888e-05,
+      "loss": 0.7012,
+      "step": 2905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.442444798637396,
+      "learning_rate": 4.9769856763214794e-05,
+      "loss": 0.9932,
+      "step": 2906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.240171769799341,
+      "learning_rate": 4.976929182523567e-05,
+      "loss": 1.5693,
+      "step": 2907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.825081552324236,
+      "learning_rate": 4.9768726197937256e-05,
+      "loss": 0.6636,
+      "step": 2908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.092352552568736,
+      "learning_rate": 4.9768159881335264e-05,
+      "loss": 1.1934,
+      "step": 2909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1602193812825017,
+      "learning_rate": 4.976759287544547e-05,
+      "loss": 1.1655,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.66670836639021,
+      "learning_rate": 4.9767025180283646e-05,
+      "loss": 0.9287,
+      "step": 2911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8773706033547874,
+      "learning_rate": 4.976645679586559e-05,
+      "loss": 0.9419,
+      "step": 2912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8829794528749753,
+      "learning_rate": 4.976588772220713e-05,
+      "loss": 0.813,
+      "step": 2913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.2137292443745675,
+      "learning_rate": 4.97653179593241e-05,
+      "loss": 0.8521,
+      "step": 2914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1363690245244977,
+      "learning_rate": 4.976474750723235e-05,
+      "loss": 0.8428,
+      "step": 2915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.41994059434598,
+      "learning_rate": 4.9764176365947745e-05,
+      "loss": 1.209,
+      "step": 2916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.8454923480368945,
+      "learning_rate": 4.976360453548621e-05,
+      "loss": 2.0664,
+      "step": 2917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2825326501598173,
+      "learning_rate": 4.976303201586364e-05,
+      "loss": 1.6035,
+      "step": 2918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.909976989550246,
+      "learning_rate": 4.976245880709596e-05,
+      "loss": 0.7041,
+      "step": 2919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9449558210135303,
+      "learning_rate": 4.976188490919914e-05,
+      "loss": 1.6016,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5578067879680035,
+      "learning_rate": 4.976131032218915e-05,
+      "loss": 0.9912,
+      "step": 2921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.866437850366236,
+      "learning_rate": 4.976073504608196e-05,
+      "loss": 1.8047,
+      "step": 2922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.153769304917711,
+      "learning_rate": 4.9760159080893604e-05,
+      "loss": 1.2969,
+      "step": 2923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.609098745896182,
+      "learning_rate": 4.9759582426640095e-05,
+      "loss": 1.6387,
+      "step": 2924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.91829600285347,
+      "learning_rate": 4.975900508333749e-05,
+      "loss": 1.584,
+      "step": 2925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3636076056635873,
+      "learning_rate": 4.9758427051001854e-05,
+      "loss": 0.8438,
+      "step": 2926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.790067234403649,
+      "learning_rate": 4.975784832964927e-05,
+      "loss": 1.3633,
+      "step": 2927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4673643448633067,
+      "learning_rate": 4.975726891929585e-05,
+      "loss": 0.7792,
+      "step": 2928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.656407789143851,
+      "learning_rate": 4.9756688819957706e-05,
+      "loss": 1.1836,
+      "step": 2929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.34836187507017,
+      "learning_rate": 4.9756108031651e-05,
+      "loss": 1.3203,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9820185208691874,
+      "learning_rate": 4.975552655439187e-05,
+      "loss": 1.8145,
+      "step": 2931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.0190698841688,
+      "learning_rate": 4.9754944388196535e-05,
+      "loss": 1.6865,
+      "step": 2932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.430235380852574,
+      "learning_rate": 4.975436153308116e-05,
+      "loss": 0.876,
+      "step": 2933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.526979152907616,
+      "learning_rate": 4.975377798906199e-05,
+      "loss": 1.4248,
+      "step": 2934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.665094476488077,
+      "learning_rate": 4.9753193756155254e-05,
+      "loss": 1.1914,
+      "step": 2935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.155369729558767,
+      "learning_rate": 4.9752608834377215e-05,
+      "loss": 1.5312,
+      "step": 2936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.190540405602983,
+      "learning_rate": 4.9752023223744136e-05,
+      "loss": 0.8643,
+      "step": 2937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.871932213761127,
+      "learning_rate": 4.975143692427234e-05,
+      "loss": 1.4385,
+      "step": 2938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3811262323850038,
+      "learning_rate": 4.975084993597813e-05,
+      "loss": 1.5781,
+      "step": 2939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4908484296061735,
+      "learning_rate": 4.975026225887783e-05,
+      "loss": 1.2617,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0520832638962947,
+      "learning_rate": 4.974967389298783e-05,
+      "loss": 0.7171,
+      "step": 2941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4611955779855657,
+      "learning_rate": 4.974908483832446e-05,
+      "loss": 1.0718,
+      "step": 2942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4567567533147185,
+      "learning_rate": 4.9748495094904146e-05,
+      "loss": 1.0332,
+      "step": 2943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8594729762440245,
+      "learning_rate": 4.974790466274328e-05,
+      "loss": 1.5605,
+      "step": 2944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.627484696136104,
+      "learning_rate": 4.974731354185831e-05,
+      "loss": 0.6748,
+      "step": 2945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.001335159630143,
+      "learning_rate": 4.9746721732265675e-05,
+      "loss": 1.4668,
+      "step": 2946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3587846104123833,
+      "learning_rate": 4.9746129233981854e-05,
+      "loss": 1.1631,
+      "step": 2947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.329085882832162,
+      "learning_rate": 4.9745536047023324e-05,
+      "loss": 1.7012,
+      "step": 2948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.205475253856122,
+      "learning_rate": 4.97449421714066e-05,
+      "loss": 1.4316,
+      "step": 2949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.285758213544695,
+      "learning_rate": 4.9744347607148215e-05,
+      "loss": 1.5508,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2307649820250917,
+      "learning_rate": 4.97437523542647e-05,
+      "loss": 1.2441,
+      "step": 2951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3554755316742653,
+      "learning_rate": 4.9743156412772635e-05,
+      "loss": 1.124,
+      "step": 2952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.7899823198065015,
+      "learning_rate": 4.97425597826886e-05,
+      "loss": 1.3613,
+      "step": 2953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.945234506381539,
+      "learning_rate": 4.97419624640292e-05,
+      "loss": 0.9385,
+      "step": 2954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.458104825922784,
+      "learning_rate": 4.9741364456811055e-05,
+      "loss": 1.4268,
+      "step": 2955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5364443382383626,
+      "learning_rate": 4.974076576105081e-05,
+      "loss": 1.1777,
+      "step": 2956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.359971826838826,
+      "learning_rate": 4.974016637676513e-05,
+      "loss": 0.7251,
+      "step": 2957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.561725348205982,
+      "learning_rate": 4.973956630397068e-05,
+      "loss": 1.043,
+      "step": 2958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3820873302551866,
+      "learning_rate": 4.973896554268419e-05,
+      "loss": 0.9717,
+      "step": 2959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.761799706603028,
+      "learning_rate": 4.973836409292234e-05,
+      "loss": 1.0337,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.9133967932679425,
+      "learning_rate": 4.97377619547019e-05,
+      "loss": 0.9072,
+      "step": 2961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.982441642148211,
+      "learning_rate": 4.973715912803961e-05,
+      "loss": 0.8926,
+      "step": 2962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0258397902606236,
+      "learning_rate": 4.973655561295226e-05,
+      "loss": 0.7559,
+      "step": 2963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.054493795409677,
+      "learning_rate": 4.9735951409456634e-05,
+      "loss": 0.8291,
+      "step": 2964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.56236424900641,
+      "learning_rate": 4.9735346517569546e-05,
+      "loss": 1.3047,
+      "step": 2965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.046111764814949,
+      "learning_rate": 4.973474093730784e-05,
+      "loss": 1.5137,
+      "step": 2966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.695576393491844,
+      "learning_rate": 4.973413466868836e-05,
+      "loss": 1.002,
+      "step": 2967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.528872583148001,
+      "learning_rate": 4.973352771172798e-05,
+      "loss": 0.9951,
+      "step": 2968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.630106400983008,
+      "learning_rate": 4.9732920066443596e-05,
+      "loss": 1.2832,
+      "step": 2969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.939177242370118,
+      "learning_rate": 4.973231173285211e-05,
+      "loss": 1.6699,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.893221951580694,
+      "learning_rate": 4.973170271097046e-05,
+      "loss": 1.3418,
+      "step": 2971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.130534426128329,
+      "learning_rate": 4.97310930008156e-05,
+      "loss": 1.3105,
+      "step": 2972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.433382224109994,
+      "learning_rate": 4.9730482602404484e-05,
+      "loss": 1.4238,
+      "step": 2973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9413205287130295,
+      "learning_rate": 4.97298715157541e-05,
+      "loss": 1.1953,
+      "step": 2974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.270919611513804,
+      "learning_rate": 4.972925974088146e-05,
+      "loss": 1.0947,
+      "step": 2975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5532739677533116,
+      "learning_rate": 4.97286472778036e-05,
+      "loss": 0.6465,
+      "step": 2976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8767479696104024,
+      "learning_rate": 4.9728034126537536e-05,
+      "loss": 1.1445,
+      "step": 2977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.06083852112335,
+      "learning_rate": 4.9727420287100356e-05,
+      "loss": 1.4453,
+      "step": 2978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5040309037539736,
+      "learning_rate": 4.972680575950914e-05,
+      "loss": 1.2168,
+      "step": 2979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.580452911427107,
+      "learning_rate": 4.972619054378098e-05,
+      "loss": 1.6938,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.229527627784632,
+      "learning_rate": 4.972557463993301e-05,
+      "loss": 1.1729,
+      "step": 2981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.382119199856342,
+      "learning_rate": 4.9724958047982364e-05,
+      "loss": 1.3213,
+      "step": 2982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7033124787721716,
+      "learning_rate": 4.9724340767946196e-05,
+      "loss": 1.6133,
+      "step": 2983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.8511139328635,
+      "learning_rate": 4.972372279984169e-05,
+      "loss": 1.3496,
+      "step": 2984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.544876995519061,
+      "learning_rate": 4.972310414368605e-05,
+      "loss": 1.1504,
+      "step": 2985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6585069070326626,
+      "learning_rate": 4.9722484799496476e-05,
+      "loss": 0.8013,
+      "step": 2986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.100456905533649,
+      "learning_rate": 4.972186476729022e-05,
+      "loss": 1.584,
+      "step": 2987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.002988175995303,
+      "learning_rate": 4.972124404708454e-05,
+      "loss": 1.4863,
+      "step": 2988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.892948515860156,
+      "learning_rate": 4.9720622638896684e-05,
+      "loss": 0.9985,
+      "step": 2989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3719624620075876,
+      "learning_rate": 4.9720000542743976e-05,
+      "loss": 1.3906,
+      "step": 2990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.161904899553502,
+      "learning_rate": 4.971937775864371e-05,
+      "loss": 2.1001,
+      "step": 2991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.101819764402736,
+      "learning_rate": 4.971875428661323e-05,
+      "loss": 1.7051,
+      "step": 2992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.289415749579306,
+      "learning_rate": 4.9718130126669884e-05,
+      "loss": 1.5312,
+      "step": 2993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8299772802729817,
+      "learning_rate": 4.9717505278831035e-05,
+      "loss": 1.0352,
+      "step": 2994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.752411130784258,
+      "learning_rate": 4.9716879743114074e-05,
+      "loss": 1.168,
+      "step": 2995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.87347611898588,
+      "learning_rate": 4.9716253519536415e-05,
+      "loss": 1.9277,
+      "step": 2996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.25323735903315,
+      "learning_rate": 4.971562660811549e-05,
+      "loss": 1.4766,
+      "step": 2997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5212057064555276,
+      "learning_rate": 4.9714999008868724e-05,
+      "loss": 1.2495,
+      "step": 2998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.712449193536176,
+      "learning_rate": 4.9714370721813604e-05,
+      "loss": 0.6504,
+      "step": 2999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3144375699965156,
+      "learning_rate": 4.97137417469676e-05,
+      "loss": 1.1602,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.867656451133663,
+      "learning_rate": 4.9713112084348235e-05,
+      "loss": 1.7891,
+      "step": 3001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.182675807565854,
+      "learning_rate": 4.971248173397301e-05,
+      "loss": 1.4688,
+      "step": 3002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.072896008250362,
+      "learning_rate": 4.97118506958595e-05,
+      "loss": 0.8535,
+      "step": 3003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8420778684732704,
+      "learning_rate": 4.971121897002523e-05,
+      "loss": 0.8896,
+      "step": 3004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1225736735489082,
+      "learning_rate": 4.971058655648779e-05,
+      "loss": 1.0664,
+      "step": 3005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.711953036414192,
+      "learning_rate": 4.97099534552648e-05,
+      "loss": 0.6064,
+      "step": 3006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.306622743335377,
+      "learning_rate": 4.970931966637386e-05,
+      "loss": 1.1484,
+      "step": 3007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6029736240148296,
+      "learning_rate": 4.970868518983261e-05,
+      "loss": 0.8613,
+      "step": 3008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7854768159278462,
+      "learning_rate": 4.970805002565872e-05,
+      "loss": 1.1372,
+      "step": 3009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.567479872674232,
+      "learning_rate": 4.970741417386984e-05,
+      "loss": 1.7852,
+      "step": 3010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2175007590268536,
+      "learning_rate": 4.970677763448369e-05,
+      "loss": 1.0625,
+      "step": 3011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1315819400694465,
+      "learning_rate": 4.970614040751798e-05,
+      "loss": 0.6851,
+      "step": 3012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.138275938098411,
+      "learning_rate": 4.970550249299043e-05,
+      "loss": 1.0513,
+      "step": 3013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.240594666723219,
+      "learning_rate": 4.970486389091881e-05,
+      "loss": 1.7969,
+      "step": 3014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.229335901044907,
+      "learning_rate": 4.970422460132089e-05,
+      "loss": 2.0684,
+      "step": 3015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.24241398920422,
+      "learning_rate": 4.970358462421445e-05,
+      "loss": 1.1445,
+      "step": 3016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.5575156865346855,
+      "learning_rate": 4.9702943959617306e-05,
+      "loss": 1.1201,
+      "step": 3017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.576529309362912,
+      "learning_rate": 4.970230260754729e-05,
+      "loss": 1.4307,
+      "step": 3018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9646746522857432,
+      "learning_rate": 4.970166056802225e-05,
+      "loss": 1.1387,
+      "step": 3019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6498238415017727,
+      "learning_rate": 4.970101784106005e-05,
+      "loss": 1.4902,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1514390595015405,
+      "learning_rate": 4.970037442667859e-05,
+      "loss": 1.1074,
+      "step": 3021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.395576485397136,
+      "learning_rate": 4.9699730324895746e-05,
+      "loss": 0.8643,
+      "step": 3022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2451129397252116,
+      "learning_rate": 4.969908553572947e-05,
+      "loss": 0.9961,
+      "step": 3023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3618311220288852,
+      "learning_rate": 4.9698440059197704e-05,
+      "loss": 1.3037,
+      "step": 3024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.601445194079998,
+      "learning_rate": 4.9697793895318404e-05,
+      "loss": 1.0537,
+      "step": 3025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4379992035869855,
+      "learning_rate": 4.969714704410956e-05,
+      "loss": 1.6113,
+      "step": 3026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.228491030059818,
+      "learning_rate": 4.969649950558915e-05,
+      "loss": 1.6035,
+      "step": 3027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.378369586873385,
+      "learning_rate": 4.969585127977523e-05,
+      "loss": 1.377,
+      "step": 3028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7936239935980915,
+      "learning_rate": 4.969520236668582e-05,
+      "loss": 0.4407,
+      "step": 3029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.962321444077485,
+      "learning_rate": 4.969455276633898e-05,
+      "loss": 0.8872,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2074396614470757,
+      "learning_rate": 4.969390247875279e-05,
+      "loss": 0.8633,
+      "step": 3031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.185273995056361,
+      "learning_rate": 4.9693251503945335e-05,
+      "loss": 0.8618,
+      "step": 3032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.695671791382751,
+      "learning_rate": 4.9692599841934754e-05,
+      "loss": 1.0,
+      "step": 3033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.660674057361059,
+      "learning_rate": 4.969194749273918e-05,
+      "loss": 1.0352,
+      "step": 3034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.686529294738708,
+      "learning_rate": 4.9691294456376744e-05,
+      "loss": 1.0488,
+      "step": 3035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9706058518428238,
+      "learning_rate": 4.969064073286564e-05,
+      "loss": 1.0195,
+      "step": 3036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.35249070339066,
+      "learning_rate": 4.968998632222406e-05,
+      "loss": 1.1152,
+      "step": 3037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.145729560951208,
+      "learning_rate": 4.9689331224470206e-05,
+      "loss": 0.874,
+      "step": 3038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.36214299809665,
+      "learning_rate": 4.9688675439622326e-05,
+      "loss": 1.0879,
+      "step": 3039
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.86159746673754,
+      "learning_rate": 4.9688018967698644e-05,
+      "loss": 0.8486,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.096939731741451,
+      "learning_rate": 4.9687361808717454e-05,
+      "loss": 1.0425,
+      "step": 3041
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.705890810505074,
+      "learning_rate": 4.968670396269703e-05,
+      "loss": 1.3555,
+      "step": 3042
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.927695861980395,
+      "learning_rate": 4.9686045429655695e-05,
+      "loss": 0.9512,
+      "step": 3043
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.500082455625593,
+      "learning_rate": 4.968538620961175e-05,
+      "loss": 1.1582,
+      "step": 3044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.892179823607114,
+      "learning_rate": 4.968472630258357e-05,
+      "loss": 1.0127,
+      "step": 3045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3621496596764295,
+      "learning_rate": 4.96840657085895e-05,
+      "loss": 1.1758,
+      "step": 3046
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.577279203307045,
+      "learning_rate": 4.968340442764793e-05,
+      "loss": 0.7046,
+      "step": 3047
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.971381404815422,
+      "learning_rate": 4.968274245977726e-05,
+      "loss": 2.0195,
+      "step": 3048
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.745468889265807,
+      "learning_rate": 4.968207980499592e-05,
+      "loss": 1.5645,
+      "step": 3049
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2437304992678437,
+      "learning_rate": 4.968141646332235e-05,
+      "loss": 0.7422,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3656567953132317,
+      "learning_rate": 4.968075243477501e-05,
+      "loss": 0.9067,
+      "step": 3051
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4324120458401555,
+      "learning_rate": 4.968008771937237e-05,
+      "loss": 1.2363,
+      "step": 3052
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.654772837798724,
+      "learning_rate": 4.967942231713294e-05,
+      "loss": 1.2617,
+      "step": 3053
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.180705389061133,
+      "learning_rate": 4.967875622807523e-05,
+      "loss": 1.4746,
+      "step": 3054
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.5786937532022,
+      "learning_rate": 4.967808945221778e-05,
+      "loss": 1.5977,
+      "step": 3055
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.248514701251255,
+      "learning_rate": 4.967742198957916e-05,
+      "loss": 1.5488,
+      "step": 3056
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.355170336999891,
+      "learning_rate": 4.967675384017791e-05,
+      "loss": 1.2285,
+      "step": 3057
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6564506903734806,
+      "learning_rate": 4.967608500403266e-05,
+      "loss": 0.9658,
+      "step": 3058
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.131521373346638,
+      "learning_rate": 4.9675415481162016e-05,
+      "loss": 1.418,
+      "step": 3059
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4742636616000038,
+      "learning_rate": 4.967474527158459e-05,
+      "loss": 1.6562,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0573902991750614,
+      "learning_rate": 4.967407437531906e-05,
+      "loss": 1.3828,
+      "step": 3061
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.688868005132546,
+      "learning_rate": 4.967340279238408e-05,
+      "loss": 1.3848,
+      "step": 3062
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.26999747673736,
+      "learning_rate": 4.9672730522798335e-05,
+      "loss": 1.5215,
+      "step": 3063
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.091545835850642,
+      "learning_rate": 4.967205756658056e-05,
+      "loss": 0.9744,
+      "step": 3064
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.042804652263804,
+      "learning_rate": 4.967138392374946e-05,
+      "loss": 1.6875,
+      "step": 3065
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5723673019067843,
+      "learning_rate": 4.9670709594323795e-05,
+      "loss": 1.0264,
+      "step": 3066
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9956511460273294,
+      "learning_rate": 4.967003457832232e-05,
+      "loss": 0.9551,
+      "step": 3067
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.935818855111339,
+      "learning_rate": 4.9669358875763825e-05,
+      "loss": 1.7324,
+      "step": 3068
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9626675586556033,
+      "learning_rate": 4.9668682486667115e-05,
+      "loss": 1.3516,
+      "step": 3069
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.801301830651444,
+      "learning_rate": 4.9668005411051015e-05,
+      "loss": 1.0693,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6460883932358334,
+      "learning_rate": 4.966732764893437e-05,
+      "loss": 1.0059,
+      "step": 3071
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.901057095193678,
+      "learning_rate": 4.9666649200336035e-05,
+      "loss": 1.2812,
+      "step": 3072
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2083562333940954,
+      "learning_rate": 4.966597006527489e-05,
+      "loss": 0.6487,
+      "step": 3073
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8365809877861676,
+      "learning_rate": 4.966529024376985e-05,
+      "loss": 0.8057,
+      "step": 3074
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.393992392790154,
+      "learning_rate": 4.966460973583982e-05,
+      "loss": 1.6758,
+      "step": 3075
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.317040290709411,
+      "learning_rate": 4.966392854150374e-05,
+      "loss": 1.8008,
+      "step": 3076
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8877448851524052,
+      "learning_rate": 4.966324666078057e-05,
+      "loss": 1.3086,
+      "step": 3077
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.316573308250614,
+      "learning_rate": 4.966256409368929e-05,
+      "loss": 1.8101,
+      "step": 3078
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.2510606713345735,
+      "learning_rate": 4.966188084024889e-05,
+      "loss": 0.9863,
+      "step": 3079
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3914364793803966,
+      "learning_rate": 4.966119690047838e-05,
+      "loss": 0.194,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.126796848102598,
+      "learning_rate": 4.96605122743968e-05,
+      "loss": 1.1777,
+      "step": 3081
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.7963182536720295,
+      "learning_rate": 4.9659826962023206e-05,
+      "loss": 1.2148,
+      "step": 3082
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.750654290658128,
+      "learning_rate": 4.965914096337667e-05,
+      "loss": 0.8965,
+      "step": 3083
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1085202215942407,
+      "learning_rate": 4.965845427847627e-05,
+      "loss": 0.6714,
+      "step": 3084
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.359739970090016,
+      "learning_rate": 4.965776690734114e-05,
+      "loss": 1.1055,
+      "step": 3085
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8209892619369956,
+      "learning_rate": 4.965707884999038e-05,
+      "loss": 0.5817,
+      "step": 3086
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0398985195789643,
+      "learning_rate": 4.965639010644316e-05,
+      "loss": 1.1099,
+      "step": 3087
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8846889271286824,
+      "learning_rate": 4.965570067671865e-05,
+      "loss": 1.1211,
+      "step": 3088
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.681013396523118,
+      "learning_rate": 4.965501056083602e-05,
+      "loss": 1.0293,
+      "step": 3089
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.26701386074007,
+      "learning_rate": 4.965431975881448e-05,
+      "loss": 1.1084,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5158614015570206,
+      "learning_rate": 4.965362827067326e-05,
+      "loss": 0.9482,
+      "step": 3091
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.647186884244474,
+      "learning_rate": 4.96529360964316e-05,
+      "loss": 1.1035,
+      "step": 3092
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.781371059529037,
+      "learning_rate": 4.965224323610877e-05,
+      "loss": 1.7285,
+      "step": 3093
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6343287431392346,
+      "learning_rate": 4.965154968972404e-05,
+      "loss": 1.0762,
+      "step": 3094
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.4225634277823,
+      "learning_rate": 4.965085545729672e-05,
+      "loss": 1.4961,
+      "step": 3095
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.311734236968757,
+      "learning_rate": 4.9650160538846133e-05,
+      "loss": 1.0996,
+      "step": 3096
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.441453905784884,
+      "learning_rate": 4.964946493439161e-05,
+      "loss": 1.125,
+      "step": 3097
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.396587139941908,
+      "learning_rate": 4.964876864395251e-05,
+      "loss": 1.0117,
+      "step": 3098
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5642472799739897,
+      "learning_rate": 4.964807166754822e-05,
+      "loss": 1.0781,
+      "step": 3099
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8662313782567106,
+      "learning_rate": 4.964737400519812e-05,
+      "loss": 0.9331,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.190973691493194,
+      "learning_rate": 4.9646675656921646e-05,
+      "loss": 1.166,
+      "step": 3101
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.748256162474036,
+      "learning_rate": 4.9645976622738216e-05,
+      "loss": 1.1797,
+      "step": 3102
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.133935206665741,
+      "learning_rate": 4.9645276902667296e-05,
+      "loss": 0.8135,
+      "step": 3103
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.131108312287856,
+      "learning_rate": 4.964457649672835e-05,
+      "loss": 0.3672,
+      "step": 3104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.2575207868246165,
+      "learning_rate": 4.9643875404940864e-05,
+      "loss": 1.9883,
+      "step": 3105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.222760590032594,
+      "learning_rate": 4.964317362732436e-05,
+      "loss": 1.9492,
+      "step": 3106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.624236677085285,
+      "learning_rate": 4.964247116389837e-05,
+      "loss": 1.0723,
+      "step": 3107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.248234904203464,
+      "learning_rate": 4.9641768014682444e-05,
+      "loss": 1.4707,
+      "step": 3108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.994987591954991,
+      "learning_rate": 4.964106417969614e-05,
+      "loss": 1.0869,
+      "step": 3109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5804282332775594,
+      "learning_rate": 4.9640359658959054e-05,
+      "loss": 0.8198,
+      "step": 3110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1820754230096817,
+      "learning_rate": 4.963965445249079e-05,
+      "loss": 0.8018,
+      "step": 3111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.1371711792794805,
+      "learning_rate": 4.9638948560310975e-05,
+      "loss": 1.2949,
+      "step": 3112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.029747261470998,
+      "learning_rate": 4.963824198243925e-05,
+      "loss": 1.0508,
+      "step": 3113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9865719363255203,
+      "learning_rate": 4.963753471889528e-05,
+      "loss": 0.5903,
+      "step": 3114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0860380798550797,
+      "learning_rate": 4.9636826769698754e-05,
+      "loss": 1.0625,
+      "step": 3115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2150508507947273,
+      "learning_rate": 4.9636118134869356e-05,
+      "loss": 1.3203,
+      "step": 3116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3653385003982126,
+      "learning_rate": 4.963540881442683e-05,
+      "loss": 1.5117,
+      "step": 3117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.647694350391192,
+      "learning_rate": 4.96346988083909e-05,
+      "loss": 1.1824,
+      "step": 3118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.404357016328504,
+      "learning_rate": 4.963398811678134e-05,
+      "loss": 1.7695,
+      "step": 3119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9480490682010914,
+      "learning_rate": 4.963327673961791e-05,
+      "loss": 0.9922,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.341942662528336,
+      "learning_rate": 4.9632564676920425e-05,
+      "loss": 1.2988,
+      "step": 3121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5780107060203423,
+      "learning_rate": 4.963185192870868e-05,
+      "loss": 0.6005,
+      "step": 3122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6470087100427837,
+      "learning_rate": 4.963113849500254e-05,
+      "loss": 0.8135,
+      "step": 3123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8326337923132563,
+      "learning_rate": 4.963042437582183e-05,
+      "loss": 1.207,
+      "step": 3124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.265547547167475,
+      "learning_rate": 4.9629709571186444e-05,
+      "loss": 1.5098,
+      "step": 3125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3117836231555895,
+      "learning_rate": 4.9628994081116265e-05,
+      "loss": 1.1523,
+      "step": 3126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.321469208078715,
+      "learning_rate": 4.9628277905631206e-05,
+      "loss": 1.0996,
+      "step": 3127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.880837255171559,
+      "learning_rate": 4.9627561044751196e-05,
+      "loss": 1.1211,
+      "step": 3128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.688354179016131,
+      "learning_rate": 4.962684349849619e-05,
+      "loss": 0.8389,
+      "step": 3129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2310435394593666,
+      "learning_rate": 4.962612526688616e-05,
+      "loss": 0.9785,
+      "step": 3130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.061034288656978,
+      "learning_rate": 4.962540634994108e-05,
+      "loss": 1.041,
+      "step": 3131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.330457606913992,
+      "learning_rate": 4.962468674768098e-05,
+      "loss": 0.8203,
+      "step": 3132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.481656872807074,
+      "learning_rate": 4.962396646012586e-05,
+      "loss": 1.3857,
+      "step": 3133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4834890565028194,
+      "learning_rate": 4.962324548729579e-05,
+      "loss": 0.7988,
+      "step": 3134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4263784561073654,
+      "learning_rate": 4.962252382921081e-05,
+      "loss": 1.0684,
+      "step": 3135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.157683387117992,
+      "learning_rate": 4.962180148589102e-05,
+      "loss": 1.1914,
+      "step": 3136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4517183583729514,
+      "learning_rate": 4.9621078457356515e-05,
+      "loss": 1.0332,
+      "step": 3137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.7439779176207875,
+      "learning_rate": 4.962035474362743e-05,
+      "loss": 1.8926,
+      "step": 3138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.034151674171983,
+      "learning_rate": 4.961963034472389e-05,
+      "loss": 0.9473,
+      "step": 3139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.521341025771267,
+      "learning_rate": 4.9618905260666055e-05,
+      "loss": 1.1445,
+      "step": 3140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.0400009346007435,
+      "learning_rate": 4.961817949147412e-05,
+      "loss": 1.2051,
+      "step": 3141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5045673610457406,
+      "learning_rate": 4.961745303716826e-05,
+      "loss": 1.3633,
+      "step": 3142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.28477255147471,
+      "learning_rate": 4.961672589776871e-05,
+      "loss": 1.0547,
+      "step": 3143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.578521839773967,
+      "learning_rate": 4.9615998073295694e-05,
+      "loss": 2.0039,
+      "step": 3144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0764766264899026,
+      "learning_rate": 4.961526956376948e-05,
+      "loss": 0.9473,
+      "step": 3145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.516907400962751,
+      "learning_rate": 4.9614540369210325e-05,
+      "loss": 1.4062,
+      "step": 3146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7571746693744643,
+      "learning_rate": 4.9613810489638545e-05,
+      "loss": 1.0693,
+      "step": 3147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.41845552684582,
+      "learning_rate": 4.961307992507443e-05,
+      "loss": 0.9834,
+      "step": 3148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9329589418213755,
+      "learning_rate": 4.9612348675538324e-05,
+      "loss": 1.2124,
+      "step": 3149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.276968933423843,
+      "learning_rate": 4.9611616741050574e-05,
+      "loss": 0.7422,
+      "step": 3150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8251100461999195,
+      "learning_rate": 4.961088412163155e-05,
+      "loss": 1.043,
+      "step": 3151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.468167103388716,
+      "learning_rate": 4.961015081730164e-05,
+      "loss": 0.7686,
+      "step": 3152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1481579407959592,
+      "learning_rate": 4.960941682808126e-05,
+      "loss": 1.0342,
+      "step": 3153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.5321777268926855,
+      "learning_rate": 4.9608682153990816e-05,
+      "loss": 0.9131,
+      "step": 3154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.397155519920435,
+      "learning_rate": 4.960794679505076e-05,
+      "loss": 0.75,
+      "step": 3155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.50265656557676,
+      "learning_rate": 4.9607210751281574e-05,
+      "loss": 1.6162,
+      "step": 3156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7587629630955703,
+      "learning_rate": 4.960647402270373e-05,
+      "loss": 1.5918,
+      "step": 3157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.653376697389661,
+      "learning_rate": 4.9605736609337725e-05,
+      "loss": 0.8604,
+      "step": 3158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.778811811204977,
+      "learning_rate": 4.96049985112041e-05,
+      "loss": 1.2617,
+      "step": 3159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.235844669824849,
+      "learning_rate": 4.960425972832337e-05,
+      "loss": 1.2305,
+      "step": 3160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4039063396545775,
+      "learning_rate": 4.960352026071612e-05,
+      "loss": 0.686,
+      "step": 3161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6328562384801155,
+      "learning_rate": 4.96027801084029e-05,
+      "loss": 1.2979,
+      "step": 3162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.463742650362446,
+      "learning_rate": 4.960203927140434e-05,
+      "loss": 0.8555,
+      "step": 3163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6347052780791036,
+      "learning_rate": 4.9601297749741036e-05,
+      "loss": 1.123,
+      "step": 3164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1483599556207436,
+      "learning_rate": 4.9600555543433636e-05,
+      "loss": 0.8232,
+      "step": 3165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.349171281120063,
+      "learning_rate": 4.959981265250278e-05,
+      "loss": 1.4473,
+      "step": 3166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.481153114406712,
+      "learning_rate": 4.9599069076969165e-05,
+      "loss": 0.824,
+      "step": 3167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0827484327050994,
+      "learning_rate": 4.959832481685347e-05,
+      "loss": 0.8872,
+      "step": 3168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.935095106985179,
+      "learning_rate": 4.9597579872176406e-05,
+      "loss": 1.3682,
+      "step": 3169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.188109994003078,
+      "learning_rate": 4.95968342429587e-05,
+      "loss": 1.2012,
+      "step": 3170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.105471178330046,
+      "learning_rate": 4.9596087929221126e-05,
+      "loss": 1.3086,
+      "step": 3171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.410165199014604,
+      "learning_rate": 4.959534093098442e-05,
+      "loss": 0.9761,
+      "step": 3172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.37901239474054,
+      "learning_rate": 4.9594593248269404e-05,
+      "loss": 2.0977,
+      "step": 3173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.660513205532945,
+      "learning_rate": 4.9593844881096865e-05,
+      "loss": 0.9453,
+      "step": 3174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.8535598264113,
+      "learning_rate": 4.959309582948763e-05,
+      "loss": 0.545,
+      "step": 3175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8135863007657487,
+      "learning_rate": 4.959234609346256e-05,
+      "loss": 1.3311,
+      "step": 3176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7715694530028654,
+      "learning_rate": 4.959159567304251e-05,
+      "loss": 0.7188,
+      "step": 3177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.300779371994658,
+      "learning_rate": 4.959084456824835e-05,
+      "loss": 1.2471,
+      "step": 3178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8722119288338983,
+      "learning_rate": 4.959009277910101e-05,
+      "loss": 0.8457,
+      "step": 3179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.828038335494476,
+      "learning_rate": 4.958934030562139e-05,
+      "loss": 1.418,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.697152131230252,
+      "learning_rate": 4.9588587147830444e-05,
+      "loss": 1.3066,
+      "step": 3181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7966891270733134,
+      "learning_rate": 4.958783330574913e-05,
+      "loss": 1.7988,
+      "step": 3182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.503313675766906,
+      "learning_rate": 4.958707877939842e-05,
+      "loss": 1.5762,
+      "step": 3183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.325175702415525,
+      "learning_rate": 4.958632356879932e-05,
+      "loss": 0.9912,
+      "step": 3184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.454501392047316,
+      "learning_rate": 4.9585567673972845e-05,
+      "loss": 1.79,
+      "step": 3185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4977594826191565,
+      "learning_rate": 4.9584811094940023e-05,
+      "loss": 0.7549,
+      "step": 3186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.467973849494497,
+      "learning_rate": 4.958405383172192e-05,
+      "loss": 1.6479,
+      "step": 3187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5114277008057613,
+      "learning_rate": 4.9583295884339615e-05,
+      "loss": 1.0928,
+      "step": 3188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.725286934121069,
+      "learning_rate": 4.9582537252814184e-05,
+      "loss": 1.165,
+      "step": 3189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3064132044412444,
+      "learning_rate": 4.958177793716675e-05,
+      "loss": 0.5671,
+      "step": 3190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5219982282555184,
+      "learning_rate": 4.958101793741845e-05,
+      "loss": 1.1738,
+      "step": 3191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.547296321882289,
+      "learning_rate": 4.958025725359042e-05,
+      "loss": 1.291,
+      "step": 3192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.032710557467225,
+      "learning_rate": 4.957949588570384e-05,
+      "loss": 1.6426,
+      "step": 3193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.522987218534232,
+      "learning_rate": 4.95787338337799e-05,
+      "loss": 0.8462,
+      "step": 3194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0120237841404305,
+      "learning_rate": 4.9577971097839806e-05,
+      "loss": 0.3513,
+      "step": 3195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.76542350551056,
+      "learning_rate": 4.957720767790477e-05,
+      "loss": 1.2305,
+      "step": 3196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.034440544890197,
+      "learning_rate": 4.957644357399606e-05,
+      "loss": 1.4258,
+      "step": 3197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1577229603677384,
+      "learning_rate": 4.9575678786134924e-05,
+      "loss": 1.5605,
+      "step": 3198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8136719699040316,
+      "learning_rate": 4.9574913314342655e-05,
+      "loss": 0.4761,
+      "step": 3199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.138597516722465,
+      "learning_rate": 4.957414715864055e-05,
+      "loss": 0.7896,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.961483706216784,
+      "learning_rate": 4.9573380319049936e-05,
+      "loss": 1.8906,
+      "step": 3201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2900689186831724,
+      "learning_rate": 4.957261279559215e-05,
+      "loss": 1.417,
+      "step": 3202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.813462990857566,
+      "learning_rate": 4.957184458828855e-05,
+      "loss": 1.5557,
+      "step": 3203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1611720316342207,
+      "learning_rate": 4.957107569716053e-05,
+      "loss": 1.3262,
+      "step": 3204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.094639051116107,
+      "learning_rate": 4.957030612222947e-05,
+      "loss": 1.1641,
+      "step": 3205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.124629495765725,
+      "learning_rate": 4.9569535863516794e-05,
+      "loss": 1.5312,
+      "step": 3206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4581650076350807,
+      "learning_rate": 4.956876492104393e-05,
+      "loss": 0.8887,
+      "step": 3207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.66361966210673,
+      "learning_rate": 4.9567993294832346e-05,
+      "loss": 1.3496,
+      "step": 3208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0947234893989695,
+      "learning_rate": 4.956722098490351e-05,
+      "loss": 0.9893,
+      "step": 3209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.175150565835053,
+      "learning_rate": 4.956644799127891e-05,
+      "loss": 1.5449,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.738686219173516,
+      "learning_rate": 4.956567431398007e-05,
+      "loss": 1.4229,
+      "step": 3211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.005221320273613,
+      "learning_rate": 4.956489995302852e-05,
+      "loss": 1.3477,
+      "step": 3212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.782761877041899,
+      "learning_rate": 4.95641249084458e-05,
+      "loss": 0.8992,
+      "step": 3213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7174609859345935,
+      "learning_rate": 4.956334918025348e-05,
+      "loss": 0.752,
+      "step": 3214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.841448481028195,
+      "learning_rate": 4.956257276847315e-05,
+      "loss": 0.8369,
+      "step": 3215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.122992213893911,
+      "learning_rate": 4.956179567312642e-05,
+      "loss": 1.416,
+      "step": 3216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.5996139364361674,
+      "learning_rate": 4.956101789423492e-05,
+      "loss": 1.8887,
+      "step": 3217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.47690136491444,
+      "learning_rate": 4.956023943182029e-05,
+      "loss": 1.585,
+      "step": 3218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3445380877554345,
+      "learning_rate": 4.95594602859042e-05,
+      "loss": 0.9648,
+      "step": 3219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7139684887495754,
+      "learning_rate": 4.955868045650832e-05,
+      "loss": 0.9639,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.719853341687948,
+      "learning_rate": 4.955789994365436e-05,
+      "loss": 1.3633,
+      "step": 3221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.185553245614192,
+      "learning_rate": 4.9557118747364054e-05,
+      "loss": 1.6445,
+      "step": 3222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.425124222238138,
+      "learning_rate": 4.955633686765911e-05,
+      "loss": 0.9795,
+      "step": 3223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.76491171092003,
+      "learning_rate": 4.9555554304561326e-05,
+      "loss": 1.4844,
+      "step": 3224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6182240725067363,
+      "learning_rate": 4.955477105809245e-05,
+      "loss": 1.0176,
+      "step": 3225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 38.05529135606916,
+      "learning_rate": 4.95539871282743e-05,
+      "loss": 1.1777,
+      "step": 3226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 29.9495139972567,
+      "learning_rate": 4.955320251512869e-05,
+      "loss": 1.0664,
+      "step": 3227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.561538804064515,
+      "learning_rate": 4.9552417218677435e-05,
+      "loss": 1.0312,
+      "step": 3228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0010976372826357,
+      "learning_rate": 4.955163123894242e-05,
+      "loss": 0.9941,
+      "step": 3229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.217034563342214,
+      "learning_rate": 4.955084457594549e-05,
+      "loss": 0.8179,
+      "step": 3230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2278834264568665,
+      "learning_rate": 4.955005722970855e-05,
+      "loss": 1.1387,
+      "step": 3231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.775722761996826,
+      "learning_rate": 4.9549269200253515e-05,
+      "loss": 0.9728,
+      "step": 3232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3252149584129627,
+      "learning_rate": 4.954848048760231e-05,
+      "loss": 1.7441,
+      "step": 3233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.295620738003973,
+      "learning_rate": 4.9547691091776896e-05,
+      "loss": 1.4873,
+      "step": 3234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4004386977743173,
+      "learning_rate": 4.9546901012799224e-05,
+      "loss": 1.2246,
+      "step": 3235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.586997959871946,
+      "learning_rate": 4.954611025069129e-05,
+      "loss": 1.4707,
+      "step": 3236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3153756093711886,
+      "learning_rate": 4.95453188054751e-05,
+      "loss": 1.0322,
+      "step": 3237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.261285943450135,
+      "learning_rate": 4.9544526677172684e-05,
+      "loss": 1.2012,
+      "step": 3238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.739158326652634,
+      "learning_rate": 4.954373386580608e-05,
+      "loss": 0.9912,
+      "step": 3239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6626952122148446,
+      "learning_rate": 4.9542940371397356e-05,
+      "loss": 0.9287,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7595022930364523,
+      "learning_rate": 4.9542146193968594e-05,
+      "loss": 1.0098,
+      "step": 3241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.688045357662075,
+      "learning_rate": 4.954135133354189e-05,
+      "loss": 0.9873,
+      "step": 3242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.99439526574354,
+      "learning_rate": 4.954055579013937e-05,
+      "loss": 1.4258,
+      "step": 3243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6649054930285394,
+      "learning_rate": 4.9539759563783176e-05,
+      "loss": 0.5991,
+      "step": 3244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4441237009763386,
+      "learning_rate": 4.953896265449546e-05,
+      "loss": 1.2063,
+      "step": 3245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6222008812928554,
+      "learning_rate": 4.95381650622984e-05,
+      "loss": 1.5879,
+      "step": 3246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9044652619863007,
+      "learning_rate": 4.9537366787214205e-05,
+      "loss": 0.6025,
+      "step": 3247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2386712871198626,
+      "learning_rate": 4.953656782926508e-05,
+      "loss": 1.2285,
+      "step": 3248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8396616074126806,
+      "learning_rate": 4.9535768188473254e-05,
+      "loss": 1.207,
+      "step": 3249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.192281151293442,
+      "learning_rate": 4.9534967864860994e-05,
+      "loss": 1.2812,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.519193796909716,
+      "learning_rate": 4.953416685845056e-05,
+      "loss": 0.9858,
+      "step": 3251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.738340409613489,
+      "learning_rate": 4.9533365169264254e-05,
+      "loss": 1.1318,
+      "step": 3252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5972041833787176,
+      "learning_rate": 4.953256279732439e-05,
+      "loss": 0.9482,
+      "step": 3253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.071380570048126,
+      "learning_rate": 4.953175974265328e-05,
+      "loss": 1.2227,
+      "step": 3254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.806703570780965,
+      "learning_rate": 4.9530956005273285e-05,
+      "loss": 1.2285,
+      "step": 3255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.313598062811762,
+      "learning_rate": 4.953015158520678e-05,
+      "loss": 1.4004,
+      "step": 3256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.309866059383876,
+      "learning_rate": 4.952934648247613e-05,
+      "loss": 1.1855,
+      "step": 3257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.23610412652668,
+      "learning_rate": 4.9528540697103767e-05,
+      "loss": 1.7773,
+      "step": 3258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.023505631499789,
+      "learning_rate": 4.952773422911209e-05,
+      "loss": 0.9844,
+      "step": 3259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.548374974684197,
+      "learning_rate": 4.952692707852356e-05,
+      "loss": 1.2158,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.064730926654741,
+      "learning_rate": 4.9526119245360636e-05,
+      "loss": 0.8525,
+      "step": 3261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1304784437143245,
+      "learning_rate": 4.952531072964579e-05,
+      "loss": 0.6609,
+      "step": 3262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.50285685973743,
+      "learning_rate": 4.9524501531401535e-05,
+      "loss": 1.165,
+      "step": 3263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0184176341680535,
+      "learning_rate": 4.9523691650650395e-05,
+      "loss": 0.8467,
+      "step": 3264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.891976498608584,
+      "learning_rate": 4.9522881087414886e-05,
+      "loss": 1.3848,
+      "step": 3265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.3766022574256915,
+      "learning_rate": 4.952206984171759e-05,
+      "loss": 1.7471,
+      "step": 3266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5292410708129642,
+      "learning_rate": 4.9521257913581065e-05,
+      "loss": 0.8418,
+      "step": 3267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3582148478189016,
+      "learning_rate": 4.952044530302792e-05,
+      "loss": 0.5449,
+      "step": 3268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3153233175891277,
+      "learning_rate": 4.951963201008076e-05,
+      "loss": 0.9736,
+      "step": 3269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1408469657841636,
+      "learning_rate": 4.9518818034762226e-05,
+      "loss": 1.3496,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8158853289870733,
+      "learning_rate": 4.9518003377094966e-05,
+      "loss": 0.6658,
+      "step": 3271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.432286140464459,
+      "learning_rate": 4.951718803710166e-05,
+      "loss": 1.1104,
+      "step": 3272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.910540355257547,
+      "learning_rate": 4.951637201480499e-05,
+      "loss": 2.2695,
+      "step": 3273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.651220292869373,
+      "learning_rate": 4.951555531022766e-05,
+      "loss": 1.3789,
+      "step": 3274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.888339071853258,
+      "learning_rate": 4.9514737923392415e-05,
+      "loss": 1.2754,
+      "step": 3275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.899545157984255,
+      "learning_rate": 4.951391985432199e-05,
+      "loss": 1.4258,
+      "step": 3276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.951038895343234,
+      "learning_rate": 4.951310110303915e-05,
+      "loss": 1.2266,
+      "step": 3277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9694343179740827,
+      "learning_rate": 4.9512281669566695e-05,
+      "loss": 1.3691,
+      "step": 3278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.259410230811874,
+      "learning_rate": 4.951146155392741e-05,
+      "loss": 1.3135,
+      "step": 3279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8282685834789651,
+      "learning_rate": 4.951064075614413e-05,
+      "loss": 0.5659,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5481646496380566,
+      "learning_rate": 4.950981927623971e-05,
+      "loss": 1.418,
+      "step": 3281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.252897092850475,
+      "learning_rate": 4.950899711423699e-05,
+      "loss": 1.1641,
+      "step": 3282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.149193301560068,
+      "learning_rate": 4.9508174270158845e-05,
+      "loss": 1.0635,
+      "step": 3283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.540988237230394,
+      "learning_rate": 4.95073507440282e-05,
+      "loss": 1.1262,
+      "step": 3284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.177444304091423,
+      "learning_rate": 4.9506526535867956e-05,
+      "loss": 1.2402,
+      "step": 3285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.030220550863021,
+      "learning_rate": 4.950570164570106e-05,
+      "loss": 1.7773,
+      "step": 3286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4177498534816912,
+      "learning_rate": 4.950487607355046e-05,
+      "loss": 1.002,
+      "step": 3287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.273226268917647,
+      "learning_rate": 4.9504049819439134e-05,
+      "loss": 1.3691,
+      "step": 3288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.837061575531963,
+      "learning_rate": 4.9503222883390086e-05,
+      "loss": 0.9795,
+      "step": 3289
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.309502592162556,
+      "learning_rate": 4.950239526542631e-05,
+      "loss": 1.9219,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.862199839897189,
+      "learning_rate": 4.950156696557085e-05,
+      "loss": 1.1758,
+      "step": 3291
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.507560741111036,
+      "learning_rate": 4.950073798384676e-05,
+      "loss": 1.0977,
+      "step": 3292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.941848944473045,
+      "learning_rate": 4.94999083202771e-05,
+      "loss": 1.623,
+      "step": 3293
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.920377992414542,
+      "learning_rate": 4.949907797488497e-05,
+      "loss": 1.0527,
+      "step": 3294
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.33720473705833,
+      "learning_rate": 4.9498246947693474e-05,
+      "loss": 1.0273,
+      "step": 3295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.197632336894078,
+      "learning_rate": 4.949741523872574e-05,
+      "loss": 0.7168,
+      "step": 3296
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8137227261550484,
+      "learning_rate": 4.949658284800491e-05,
+      "loss": 1.1104,
+      "step": 3297
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.293809986629394,
+      "learning_rate": 4.949574977555415e-05,
+      "loss": 1.0703,
+      "step": 3298
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.180313536104144,
+      "learning_rate": 4.9494916021396645e-05,
+      "loss": 0.9746,
+      "step": 3299
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3679852525293046,
+      "learning_rate": 4.949408158555561e-05,
+      "loss": 0.7539,
+      "step": 3300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.076868791869339,
+      "learning_rate": 4.949324646805425e-05,
+      "loss": 1.0938,
+      "step": 3301
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1323593832077306,
+      "learning_rate": 4.94924106689158e-05,
+      "loss": 0.7456,
+      "step": 3302
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5973771077069068,
+      "learning_rate": 4.9491574188163546e-05,
+      "loss": 0.8174,
+      "step": 3303
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.328117216980489,
+      "learning_rate": 4.949073702582075e-05,
+      "loss": 1.7559,
+      "step": 3304
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4290710844817975,
+      "learning_rate": 4.948989918191071e-05,
+      "loss": 1.0565,
+      "step": 3305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3354611118340025,
+      "learning_rate": 4.9489060656456745e-05,
+      "loss": 0.6943,
+      "step": 3306
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.851674700308092,
+      "learning_rate": 4.9488221449482197e-05,
+      "loss": 1.5332,
+      "step": 3307
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.927114826662829,
+      "learning_rate": 4.948738156101042e-05,
+      "loss": 1.4814,
+      "step": 3308
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.1905962970805755,
+      "learning_rate": 4.9486540991064765e-05,
+      "loss": 1.1426,
+      "step": 3309
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.722077098749568,
+      "learning_rate": 4.948569973966866e-05,
+      "loss": 0.9331,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.13867310127676,
+      "learning_rate": 4.948485780684548e-05,
+      "loss": 0.7925,
+      "step": 3311
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.617647805806037,
+      "learning_rate": 4.948401519261869e-05,
+      "loss": 1.5352,
+      "step": 3312
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9964237425599751,
+      "learning_rate": 4.948317189701173e-05,
+      "loss": 0.9062,
+      "step": 3313
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8815607365585856,
+      "learning_rate": 4.9482327920048054e-05,
+      "loss": 1.043,
+      "step": 3314
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7820284322002893,
+      "learning_rate": 4.948148326175116e-05,
+      "loss": 1.0732,
+      "step": 3315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.763388036164217,
+      "learning_rate": 4.948063792214455e-05,
+      "loss": 1.1865,
+      "step": 3316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0711715875276955,
+      "learning_rate": 4.947979190125176e-05,
+      "loss": 1.418,
+      "step": 3317
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6773524662406905,
+      "learning_rate": 4.9478945199096315e-05,
+      "loss": 1.0474,
+      "step": 3318
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.738909726797623,
+      "learning_rate": 4.94780978157018e-05,
+      "loss": 0.7659,
+      "step": 3319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7944232274592586,
+      "learning_rate": 4.947724975109179e-05,
+      "loss": 1.3608,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.735954684764821,
+      "learning_rate": 4.947640100528988e-05,
+      "loss": 1.248,
+      "step": 3321
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.045185352683481,
+      "learning_rate": 4.947555157831969e-05,
+      "loss": 1.5106,
+      "step": 3322
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.050680434823923,
+      "learning_rate": 4.947470147020487e-05,
+      "loss": 1.4668,
+      "step": 3323
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7904415177409163,
+      "learning_rate": 4.947385068096907e-05,
+      "loss": 1.0811,
+      "step": 3324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.291038935836224,
+      "learning_rate": 4.9472999210635975e-05,
+      "loss": 1.334,
+      "step": 3325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.642008257383984,
+      "learning_rate": 4.9472147059229265e-05,
+      "loss": 0.6343,
+      "step": 3326
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.770743509619729,
+      "learning_rate": 4.947129422677267e-05,
+      "loss": 1.2031,
+      "step": 3327
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.553696770714523,
+      "learning_rate": 4.947044071328991e-05,
+      "loss": 1.4004,
+      "step": 3328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0912342053544757,
+      "learning_rate": 4.946958651880475e-05,
+      "loss": 1.4443,
+      "step": 3329
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.308619204296411,
+      "learning_rate": 4.946873164334097e-05,
+      "loss": 1.2344,
+      "step": 3330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.27648260412308,
+      "learning_rate": 4.946787608692235e-05,
+      "loss": 0.9895,
+      "step": 3331
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4808012421413093,
+      "learning_rate": 4.946701984957269e-05,
+      "loss": 0.5125,
+      "step": 3332
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.5409151107820245,
+      "learning_rate": 4.946616293131583e-05,
+      "loss": 0.7402,
+      "step": 3333
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.467990903401729,
+      "learning_rate": 4.9465305332175614e-05,
+      "loss": 0.3302,
+      "step": 3334
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.549746188017927,
+      "learning_rate": 4.946444705217591e-05,
+      "loss": 1.3242,
+      "step": 3335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.431518747730057,
+      "learning_rate": 4.946358809134061e-05,
+      "loss": 0.9404,
+      "step": 3336
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.662308208329565,
+      "learning_rate": 4.9462728449693607e-05,
+      "loss": 1.5371,
+      "step": 3337
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.778946083004288,
+      "learning_rate": 4.9461868127258834e-05,
+      "loss": 1.9238,
+      "step": 3338
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7653596391696313,
+      "learning_rate": 4.9461007124060225e-05,
+      "loss": 0.9121,
+      "step": 3339
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1267809561342865,
+      "learning_rate": 4.946014544012175e-05,
+      "loss": 1.1055,
+      "step": 3340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2376683645997275,
+      "learning_rate": 4.9459283075467386e-05,
+      "loss": 0.7212,
+      "step": 3341
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.108728154221345,
+      "learning_rate": 4.945842003012113e-05,
+      "loss": 0.5876,
+      "step": 3342
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.382946100765648,
+      "learning_rate": 4.9457556304107e-05,
+      "loss": 1.0381,
+      "step": 3343
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.941735345283452,
+      "learning_rate": 4.945669189744903e-05,
+      "loss": 0.6163,
+      "step": 3344
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2244611494804976,
+      "learning_rate": 4.945582681017129e-05,
+      "loss": 1.0254,
+      "step": 3345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.778083865154174,
+      "learning_rate": 4.9454961042297844e-05,
+      "loss": 0.9609,
+      "step": 3346
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.294002787201709,
+      "learning_rate": 4.945409459385278e-05,
+      "loss": 1.0625,
+      "step": 3347
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.033224256521562,
+      "learning_rate": 4.945322746486023e-05,
+      "loss": 1.1973,
+      "step": 3348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8317481508975426,
+      "learning_rate": 4.9452359655344305e-05,
+      "loss": 1.0469,
+      "step": 3349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0148811015760844,
+      "learning_rate": 4.9451491165329164e-05,
+      "loss": 1.3203,
+      "step": 3350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8265687654593443,
+      "learning_rate": 4.9450621994838985e-05,
+      "loss": 1.1641,
+      "step": 3351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3440360339783317,
+      "learning_rate": 4.944975214389795e-05,
+      "loss": 0.9951,
+      "step": 3352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.7839289576865065,
+      "learning_rate": 4.9448881612530265e-05,
+      "loss": 1.5566,
+      "step": 3353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.51364142837403,
+      "learning_rate": 4.944801040076016e-05,
+      "loss": 0.7212,
+      "step": 3354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9820877990239922,
+      "learning_rate": 4.944713850861187e-05,
+      "loss": 0.6958,
+      "step": 3355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.741877851636446,
+      "learning_rate": 4.944626593610968e-05,
+      "loss": 1.0117,
+      "step": 3356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.184605197381174,
+      "learning_rate": 4.9445392683277854e-05,
+      "loss": 1.8691,
+      "step": 3357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9869952991573654,
+      "learning_rate": 4.94445187501407e-05,
+      "loss": 1.4043,
+      "step": 3358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.878009547797636,
+      "learning_rate": 4.9443644136722545e-05,
+      "loss": 0.7627,
+      "step": 3359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.9148509861805225,
+      "learning_rate": 4.944276884304772e-05,
+      "loss": 1.6191,
+      "step": 3360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2876581564671827,
+      "learning_rate": 4.944189286914059e-05,
+      "loss": 0.876,
+      "step": 3361
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3588831306735027,
+      "learning_rate": 4.9441016215025534e-05,
+      "loss": 0.6241,
+      "step": 3362
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.361307137556312,
+      "learning_rate": 4.944013888072694e-05,
+      "loss": 1.209,
+      "step": 3363
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9017161248495895,
+      "learning_rate": 4.9439260866269245e-05,
+      "loss": 1.2314,
+      "step": 3364
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8995622356448374,
+      "learning_rate": 4.9438382171676856e-05,
+      "loss": 0.624,
+      "step": 3365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.2162470421553095,
+      "learning_rate": 4.943750279697424e-05,
+      "loss": 1.0088,
+      "step": 3366
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.134166581127677,
+      "learning_rate": 4.943662274218588e-05,
+      "loss": 1.1836,
+      "step": 3367
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.358583743196144,
+      "learning_rate": 4.943574200733625e-05,
+      "loss": 1.46,
+      "step": 3368
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.1013017860032726,
+      "learning_rate": 4.9434860592449864e-05,
+      "loss": 1.0972,
+      "step": 3369
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.468113883862234,
+      "learning_rate": 4.943397849755127e-05,
+      "loss": 1.1699,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.527002043227716,
+      "learning_rate": 4.943309572266498e-05,
+      "loss": 0.6982,
+      "step": 3371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8885968261451314,
+      "learning_rate": 4.94322122678156e-05,
+      "loss": 0.6299,
+      "step": 3372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.305959614511091,
+      "learning_rate": 4.943132813302769e-05,
+      "loss": 1.1914,
+      "step": 3373
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.311482262480769,
+      "learning_rate": 4.943044331832587e-05,
+      "loss": 1.3721,
+      "step": 3374
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.12436355732844,
+      "learning_rate": 4.942955782373475e-05,
+      "loss": 1.084,
+      "step": 3375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.527038461444566,
+      "learning_rate": 4.942867164927899e-05,
+      "loss": 1.125,
+      "step": 3376
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6032738581244512,
+      "learning_rate": 4.942778479498324e-05,
+      "loss": 1.459,
+      "step": 3377
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.460581532724939,
+      "learning_rate": 4.9426897260872186e-05,
+      "loss": 0.8066,
+      "step": 3378
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.07478283262515,
+      "learning_rate": 4.942600904697052e-05,
+      "loss": 0.9238,
+      "step": 3379
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0319158783118074,
+      "learning_rate": 4.9425120153302965e-05,
+      "loss": 1.0657,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.166085519858253,
+      "learning_rate": 4.942423057989426e-05,
+      "loss": 1.2324,
+      "step": 3381
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2790308570731233,
+      "learning_rate": 4.942334032676917e-05,
+      "loss": 1.541,
+      "step": 3382
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.987277848297575,
+      "learning_rate": 4.942244939395245e-05,
+      "loss": 1.7207,
+      "step": 3383
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4321674802608833,
+      "learning_rate": 4.942155778146892e-05,
+      "loss": 1.1079,
+      "step": 3384
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.484877340601894,
+      "learning_rate": 4.9420665489343365e-05,
+      "loss": 1.25,
+      "step": 3385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.467789921331725,
+      "learning_rate": 4.941977251760064e-05,
+      "loss": 1.8965,
+      "step": 3386
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.82013127601675,
+      "learning_rate": 4.9418878866265584e-05,
+      "loss": 0.4932,
+      "step": 3387
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.109584419088407,
+      "learning_rate": 4.9417984535363064e-05,
+      "loss": 0.7124,
+      "step": 3388
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.038521871661965,
+      "learning_rate": 4.941708952491798e-05,
+      "loss": 0.7656,
+      "step": 3389
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1251339692963254,
+      "learning_rate": 4.941619383495524e-05,
+      "loss": 0.7681,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.353452206848013,
+      "learning_rate": 4.941529746549976e-05,
+      "loss": 1.3867,
+      "step": 3391
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5898147512340004,
+      "learning_rate": 4.941440041657649e-05,
+      "loss": 1.0449,
+      "step": 3392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.199029348561811,
+      "learning_rate": 4.94135026882104e-05,
+      "loss": 0.9883,
+      "step": 3393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.309358929504515,
+      "learning_rate": 4.9412604280426465e-05,
+      "loss": 1.2207,
+      "step": 3394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.114824954605711,
+      "learning_rate": 4.941170519324969e-05,
+      "loss": 0.4395,
+      "step": 3395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.301134066254125,
+      "learning_rate": 4.94108054267051e-05,
+      "loss": 0.7793,
+      "step": 3396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4780632294604708,
+      "learning_rate": 4.9409904980817736e-05,
+      "loss": 0.5771,
+      "step": 3397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.830917274253049,
+      "learning_rate": 4.940900385561265e-05,
+      "loss": 1.5078,
+      "step": 3398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.435881163640236,
+      "learning_rate": 4.9408102051114926e-05,
+      "loss": 1.2627,
+      "step": 3399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7589709736761723,
+      "learning_rate": 4.940719956734965e-05,
+      "loss": 1.3418,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7850944792162664,
+      "learning_rate": 4.9406296404341956e-05,
+      "loss": 1.0691,
+      "step": 3401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.806813237588939,
+      "learning_rate": 4.940539256211696e-05,
+      "loss": 1.4844,
+      "step": 3402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3591230649522292,
+      "learning_rate": 4.9404488040699836e-05,
+      "loss": 1.3359,
+      "step": 3403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6030562715130905,
+      "learning_rate": 4.940358284011574e-05,
+      "loss": 1.0051,
+      "step": 3404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4522462473648203,
+      "learning_rate": 4.940267696038986e-05,
+      "loss": 0.9277,
+      "step": 3405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.382959155996638,
+      "learning_rate": 4.940177040154742e-05,
+      "loss": 1.0898,
+      "step": 3406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9578517647226263,
+      "learning_rate": 4.9400863163613645e-05,
+      "loss": 1.1309,
+      "step": 3407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.760547474691047,
+      "learning_rate": 4.939995524661378e-05,
+      "loss": 2.0273,
+      "step": 3408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.251237365307136,
+      "learning_rate": 4.93990466505731e-05,
+      "loss": 1.2969,
+      "step": 3409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5757841638392964,
+      "learning_rate": 4.9398137375516884e-05,
+      "loss": 1.0544,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7805805621455058,
+      "learning_rate": 4.9397227421470424e-05,
+      "loss": 1.127,
+      "step": 3411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.028003793575833,
+      "learning_rate": 4.939631678845907e-05,
+      "loss": 1.1572,
+      "step": 3412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1544921814535476,
+      "learning_rate": 4.939540547650815e-05,
+      "loss": 1.6113,
+      "step": 3413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.939124117898026,
+      "learning_rate": 4.9394493485643024e-05,
+      "loss": 1.1504,
+      "step": 3414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.85755770941237,
+      "learning_rate": 4.939358081588908e-05,
+      "loss": 1.2383,
+      "step": 3415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.145547378040127,
+      "learning_rate": 4.9392667467271704e-05,
+      "loss": 0.9102,
+      "step": 3416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.055446364638499,
+      "learning_rate": 4.939175343981633e-05,
+      "loss": 1.207,
+      "step": 3417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.468911761577272,
+      "learning_rate": 4.9390838733548387e-05,
+      "loss": 1.0459,
+      "step": 3418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.533517165985428,
+      "learning_rate": 4.9389923348493325e-05,
+      "loss": 1.7939,
+      "step": 3419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.925445255879762,
+      "learning_rate": 4.938900728467664e-05,
+      "loss": 1.374,
+      "step": 3420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.747298227877688,
+      "learning_rate": 4.9388090542123795e-05,
+      "loss": 1.2852,
+      "step": 3421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.313386109813488,
+      "learning_rate": 4.938717312086032e-05,
+      "loss": 2.1367,
+      "step": 3422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.770353978936006,
+      "learning_rate": 4.9386255020911755e-05,
+      "loss": 1.2153,
+      "step": 3423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.204747379149893,
+      "learning_rate": 4.938533624230363e-05,
+      "loss": 1.0881,
+      "step": 3424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.112989818059217,
+      "learning_rate": 4.9384416785061535e-05,
+      "loss": 1.1919,
+      "step": 3425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.738138302378664,
+      "learning_rate": 4.938349664921103e-05,
+      "loss": 1.4355,
+      "step": 3426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3830943456839697,
+      "learning_rate": 4.938257583477776e-05,
+      "loss": 0.8232,
+      "step": 3427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.363753039974878,
+      "learning_rate": 4.9381654341787324e-05,
+      "loss": 0.7344,
+      "step": 3428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.355973335669418,
+      "learning_rate": 4.938073217026537e-05,
+      "loss": 0.7988,
+      "step": 3429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.769808895161707,
+      "learning_rate": 4.9379809320237566e-05,
+      "loss": 1.8418,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9516690554719545,
+      "learning_rate": 4.93788857917296e-05,
+      "loss": 1.0527,
+      "step": 3431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.371854375823887,
+      "learning_rate": 4.937796158476716e-05,
+      "loss": 1.2959,
+      "step": 3432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8912688798749695,
+      "learning_rate": 4.937703669937598e-05,
+      "loss": 0.728,
+      "step": 3433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.358514516260345,
+      "learning_rate": 4.937611113558178e-05,
+      "loss": 1.001,
+      "step": 3434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.9575532335060375,
+      "learning_rate": 4.937518489341034e-05,
+      "loss": 1.7012,
+      "step": 3435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8414238439124837,
+      "learning_rate": 4.937425797288743e-05,
+      "loss": 1.042,
+      "step": 3436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.332378850923331,
+      "learning_rate": 4.937333037403884e-05,
+      "loss": 0.9688,
+      "step": 3437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1522517799471044,
+      "learning_rate": 4.937240209689039e-05,
+      "loss": 1.7422,
+      "step": 3438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7080010332530993,
+      "learning_rate": 4.9371473141467905e-05,
+      "loss": 1.3965,
+      "step": 3439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.407331743708702,
+      "learning_rate": 4.937054350779725e-05,
+      "loss": 0.9722,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.421323708180564,
+      "learning_rate": 4.9369613195904285e-05,
+      "loss": 1.2588,
+      "step": 3441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.572459704183197,
+      "learning_rate": 4.9368682205814916e-05,
+      "loss": 0.7026,
+      "step": 3442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.054814788008593,
+      "learning_rate": 4.9367750537555035e-05,
+      "loss": 1.3555,
+      "step": 3443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.8014417895173445,
+      "learning_rate": 4.9366818191150575e-05,
+      "loss": 1.1465,
+      "step": 3444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.096389180412212,
+      "learning_rate": 4.936588516662749e-05,
+      "loss": 2.2285,
+      "step": 3445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.140131973360296,
+      "learning_rate": 4.9364951464011735e-05,
+      "loss": 1.4707,
+      "step": 3446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.160120717644618,
+      "learning_rate": 4.9364017083329296e-05,
+      "loss": 1.5098,
+      "step": 3447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.608796929383774,
+      "learning_rate": 4.936308202460619e-05,
+      "loss": 0.9951,
+      "step": 3448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.955124300463476,
+      "learning_rate": 4.936214628786843e-05,
+      "loss": 0.916,
+      "step": 3449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5460522743584746,
+      "learning_rate": 4.936120987314204e-05,
+      "loss": 0.7119,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.269353892559401,
+      "learning_rate": 4.936027278045311e-05,
+      "loss": 0.5859,
+      "step": 3451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3068590181924153,
+      "learning_rate": 4.9359335009827706e-05,
+      "loss": 1.457,
+      "step": 3452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0109086388163337,
+      "learning_rate": 4.935839656129192e-05,
+      "loss": 2.0488,
+      "step": 3453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5104765089841288,
+      "learning_rate": 4.935745743487188e-05,
+      "loss": 0.9775,
+      "step": 3454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.692936913966594,
+      "learning_rate": 4.9356517630593716e-05,
+      "loss": 1.1484,
+      "step": 3455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8161595330807863,
+      "learning_rate": 4.935557714848358e-05,
+      "loss": 0.5014,
+      "step": 3456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.756286877603216,
+      "learning_rate": 4.9354635988567646e-05,
+      "loss": 1.2578,
+      "step": 3457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0279018097079784,
+      "learning_rate": 4.93536941508721e-05,
+      "loss": 0.7212,
+      "step": 3458
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.251280855425521,
+      "learning_rate": 4.9352751635423164e-05,
+      "loss": 1.2266,
+      "step": 3459
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.57280727766352,
+      "learning_rate": 4.9351808442247075e-05,
+      "loss": 1.5586,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.955230825479347,
+      "learning_rate": 4.9350864571370056e-05,
+      "loss": 1.6035,
+      "step": 3461
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.3890786485385105,
+      "learning_rate": 4.93499200228184e-05,
+      "loss": 2.1445,
+      "step": 3462
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.02473283503477,
+      "learning_rate": 4.934897479661838e-05,
+      "loss": 1.5508,
+      "step": 3463
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6136758047452617,
+      "learning_rate": 4.934802889279631e-05,
+      "loss": 1.062,
+      "step": 3464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9412929315510032,
+      "learning_rate": 4.934708231137849e-05,
+      "loss": 1.1138,
+      "step": 3465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.941321342949337,
+      "learning_rate": 4.934613505239129e-05,
+      "loss": 0.7344,
+      "step": 3466
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.87606599973186,
+      "learning_rate": 4.934518711586107e-05,
+      "loss": 0.5432,
+      "step": 3467
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6705823855116133,
+      "learning_rate": 4.934423850181419e-05,
+      "loss": 1.0762,
+      "step": 3468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.253748949402931,
+      "learning_rate": 4.934328921027708e-05,
+      "loss": 0.8135,
+      "step": 3469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.354001291919694,
+      "learning_rate": 4.9342339241276126e-05,
+      "loss": 2.168,
+      "step": 3470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2320196368403042,
+      "learning_rate": 4.9341388594837787e-05,
+      "loss": 0.9585,
+      "step": 3471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.880782653359512,
+      "learning_rate": 4.9340437270988514e-05,
+      "loss": 0.9663,
+      "step": 3472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.0061321409606,
+      "learning_rate": 4.933948526975478e-05,
+      "loss": 1.8027,
+      "step": 3473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.558937165055322,
+      "learning_rate": 4.933853259116308e-05,
+      "loss": 0.9023,
+      "step": 3474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5073255975379607,
+      "learning_rate": 4.933757923523993e-05,
+      "loss": 0.7324,
+      "step": 3475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3964211982875088,
+      "learning_rate": 4.933662520201184e-05,
+      "loss": 0.7576,
+      "step": 3476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.110973256415463,
+      "learning_rate": 4.9335670491505397e-05,
+      "loss": 1.5801,
+      "step": 3477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.541963377846717,
+      "learning_rate": 4.9334715103747145e-05,
+      "loss": 1.6445,
+      "step": 3478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.364117031640869,
+      "learning_rate": 4.933375903876368e-05,
+      "loss": 1.4238,
+      "step": 3479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6079535353037766,
+      "learning_rate": 4.93328022965816e-05,
+      "loss": 1.2363,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4820590950926174,
+      "learning_rate": 4.933184487722755e-05,
+      "loss": 0.8428,
+      "step": 3481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9800736490460897,
+      "learning_rate": 4.933088678072816e-05,
+      "loss": 0.9727,
+      "step": 3482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.616343438985535,
+      "learning_rate": 4.932992800711009e-05,
+      "loss": 1.3262,
+      "step": 3483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.700066572710812,
+      "learning_rate": 4.9328968556400026e-05,
+      "loss": 1.0093,
+      "step": 3484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9115584918669417,
+      "learning_rate": 4.932800842862469e-05,
+      "loss": 0.9521,
+      "step": 3485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.335070784008938,
+      "learning_rate": 4.932704762381076e-05,
+      "loss": 1.3301,
+      "step": 3486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1570729330056655,
+      "learning_rate": 4.9326086141985006e-05,
+      "loss": 1.3867,
+      "step": 3487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.679930855059978,
+      "learning_rate": 4.9325123983174184e-05,
+      "loss": 1.3418,
+      "step": 3488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6102487163498527,
+      "learning_rate": 4.9324161147405055e-05,
+      "loss": 1.1455,
+      "step": 3489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9112522902318942,
+      "learning_rate": 4.9323197634704434e-05,
+      "loss": 0.7539,
+      "step": 3490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.80515817133707,
+      "learning_rate": 4.9322233445099114e-05,
+      "loss": 2.1367,
+      "step": 3491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0970968569688297,
+      "learning_rate": 4.932126857861594e-05,
+      "loss": 1.0215,
+      "step": 3492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.338525773456717,
+      "learning_rate": 4.9320303035281766e-05,
+      "loss": 0.6157,
+      "step": 3493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.097198414010163,
+      "learning_rate": 4.931933681512346e-05,
+      "loss": 1.0635,
+      "step": 3494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.833108294656759,
+      "learning_rate": 4.931836991816792e-05,
+      "loss": 0.9937,
+      "step": 3495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.159956537564022,
+      "learning_rate": 4.931740234444203e-05,
+      "loss": 1.0615,
+      "step": 3496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5488561869863267,
+      "learning_rate": 4.931643409397274e-05,
+      "loss": 0.7158,
+      "step": 3497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2658573049755866,
+      "learning_rate": 4.931546516678699e-05,
+      "loss": 1.0879,
+      "step": 3498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6102534659948793,
+      "learning_rate": 4.9314495562911734e-05,
+      "loss": 1.459,
+      "step": 3499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.414149113061162,
+      "learning_rate": 4.9313525282373974e-05,
+      "loss": 1.0918,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0135232983972284,
+      "learning_rate": 4.93125543252007e-05,
+      "loss": 0.7524,
+      "step": 3501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.613372653331152,
+      "learning_rate": 4.9311582691418936e-05,
+      "loss": 1.5352,
+      "step": 3502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.094516251858865,
+      "learning_rate": 4.9310610381055724e-05,
+      "loss": 1.6426,
+      "step": 3503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1936837123843693,
+      "learning_rate": 4.930963739413812e-05,
+      "loss": 0.9648,
+      "step": 3504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.295378216762872,
+      "learning_rate": 4.9308663730693206e-05,
+      "loss": 0.7297,
+      "step": 3505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.210339559080852,
+      "learning_rate": 4.930768939074808e-05,
+      "loss": 1.6113,
+      "step": 3506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.216244747286885,
+      "learning_rate": 4.930671437432985e-05,
+      "loss": 0.6455,
+      "step": 3507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.859900799804199,
+      "learning_rate": 4.9305738681465655e-05,
+      "loss": 1.8242,
+      "step": 3508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1169361596372798,
+      "learning_rate": 4.9304762312182645e-05,
+      "loss": 1.4707,
+      "step": 3509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9006776609861937,
+      "learning_rate": 4.930378526650799e-05,
+      "loss": 1.7578,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.622355991478831,
+      "learning_rate": 4.93028075444689e-05,
+      "loss": 0.8081,
+      "step": 3511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.769200745138608,
+      "learning_rate": 4.930182914609256e-05,
+      "loss": 1.3447,
+      "step": 3512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0639510396798557,
+      "learning_rate": 4.930085007140621e-05,
+      "loss": 1.3193,
+      "step": 3513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.618466639254392,
+      "learning_rate": 4.9299870320437094e-05,
+      "loss": 1.0859,
+      "step": 3514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.48082638062244,
+      "learning_rate": 4.929888989321247e-05,
+      "loss": 0.79,
+      "step": 3515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.547532260080312,
+      "learning_rate": 4.929790878975965e-05,
+      "loss": 1.0625,
+      "step": 3516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.578416396548409,
+      "learning_rate": 4.929692701010591e-05,
+      "loss": 1.2578,
+      "step": 3517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7755099438032347,
+      "learning_rate": 4.9295944554278585e-05,
+      "loss": 0.959,
+      "step": 3518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.497909148474792,
+      "learning_rate": 4.929496142230502e-05,
+      "loss": 1.4238,
+      "step": 3519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.280007972712919,
+      "learning_rate": 4.929397761421256e-05,
+      "loss": 1.582,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.762751078557823,
+      "learning_rate": 4.9292993130028595e-05,
+      "loss": 1.2031,
+      "step": 3521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.18389988699818,
+      "learning_rate": 4.929200796978052e-05,
+      "loss": 1.1816,
+      "step": 3522
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.81210697394935,
+      "learning_rate": 4.929102213349576e-05,
+      "loss": 0.9199,
+      "step": 3523
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6356613186348223,
+      "learning_rate": 4.929003562120174e-05,
+      "loss": 1.0947,
+      "step": 3524
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0670740651683044,
+      "learning_rate": 4.928904843292591e-05,
+      "loss": 1.5273,
+      "step": 3525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.996216677778156,
+      "learning_rate": 4.928806056869575e-05,
+      "loss": 1.084,
+      "step": 3526
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.202640389602993,
+      "learning_rate": 4.9287072028538764e-05,
+      "loss": 1.7051,
+      "step": 3527
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.270416586787175,
+      "learning_rate": 4.928608281248244e-05,
+      "loss": 0.9199,
+      "step": 3528
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0246414667788084,
+      "learning_rate": 4.928509292055432e-05,
+      "loss": 0.4189,
+      "step": 3529
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.530006797503013,
+      "learning_rate": 4.9284102352781955e-05,
+      "loss": 1.0762,
+      "step": 3530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4451426769104962,
+      "learning_rate": 4.92831111091929e-05,
+      "loss": 1.2598,
+      "step": 3531
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1684554369962896,
+      "learning_rate": 4.9282119189814756e-05,
+      "loss": 0.7686,
+      "step": 3532
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.826051399399706,
+      "learning_rate": 4.9281126594675116e-05,
+      "loss": 1.4668,
+      "step": 3533
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9636179685686628,
+      "learning_rate": 4.928013332380161e-05,
+      "loss": 1.5332,
+      "step": 3534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6934231978082197,
+      "learning_rate": 4.9279139377221875e-05,
+      "loss": 0.9951,
+      "step": 3535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3891855531266897,
+      "learning_rate": 4.927814475496358e-05,
+      "loss": 0.8252,
+      "step": 3536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.517983692263681,
+      "learning_rate": 4.9277149457054394e-05,
+      "loss": 1.166,
+      "step": 3537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.74211208528636,
+      "learning_rate": 4.927615348352203e-05,
+      "loss": 1.3789,
+      "step": 3538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.0077189830555175,
+      "learning_rate": 4.9275156834394184e-05,
+      "loss": 1.0781,
+      "step": 3539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5318696947119883,
+      "learning_rate": 4.927415950969862e-05,
+      "loss": 0.9062,
+      "step": 3540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9787357128830012,
+      "learning_rate": 4.927316150946307e-05,
+      "loss": 0.8425,
+      "step": 3541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.242041453553618,
+      "learning_rate": 4.927216283371532e-05,
+      "loss": 1.4707,
+      "step": 3542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.314512342903259,
+      "learning_rate": 4.927116348248316e-05,
+      "loss": 1.0059,
+      "step": 3543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.340475516136001,
+      "learning_rate": 4.9270163455794406e-05,
+      "loss": 1.2305,
+      "step": 3544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.566094087750389,
+      "learning_rate": 4.9269162753676874e-05,
+      "loss": 0.5862,
+      "step": 3545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.524973132649109,
+      "learning_rate": 4.9268161376158425e-05,
+      "loss": 1.0605,
+      "step": 3546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.144141656893084,
+      "learning_rate": 4.926715932326693e-05,
+      "loss": 1.1055,
+      "step": 3547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9462888208364193,
+      "learning_rate": 4.926615659503026e-05,
+      "loss": 0.8877,
+      "step": 3548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.35296122517003,
+      "learning_rate": 4.926515319147634e-05,
+      "loss": 1.0977,
+      "step": 3549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4229135563033113,
+      "learning_rate": 4.9264149112633084e-05,
+      "loss": 1.3096,
+      "step": 3550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7083029472936162,
+      "learning_rate": 4.926314435852843e-05,
+      "loss": 0.4703,
+      "step": 3551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.603548970882299,
+      "learning_rate": 4.926213892919035e-05,
+      "loss": 0.9883,
+      "step": 3552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.885237195364612,
+      "learning_rate": 4.9261132824646825e-05,
+      "loss": 1.5879,
+      "step": 3553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4467806293922045,
+      "learning_rate": 4.9260126044925846e-05,
+      "loss": 0.8584,
+      "step": 3554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.008324868680727,
+      "learning_rate": 4.925911859005543e-05,
+      "loss": 1.2119,
+      "step": 3555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.579970283189303,
+      "learning_rate": 4.925811046006363e-05,
+      "loss": 1.3496,
+      "step": 3556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6061607254070216,
+      "learning_rate": 4.925710165497849e-05,
+      "loss": 1.1074,
+      "step": 3557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8811526124402809,
+      "learning_rate": 4.925609217482808e-05,
+      "loss": 1.6348,
+      "step": 3558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7539913516050802,
+      "learning_rate": 4.92550820196405e-05,
+      "loss": 1.0117,
+      "step": 3559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9273499789348594,
+      "learning_rate": 4.925407118944386e-05,
+      "loss": 1.0605,
+      "step": 3560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8731181589742993,
+      "learning_rate": 4.92530596842663e-05,
+      "loss": 1.5762,
+      "step": 3561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2803384710346113,
+      "learning_rate": 4.9252047504135955e-05,
+      "loss": 1.8223,
+      "step": 3562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.825187505527736,
+      "learning_rate": 4.925103464908101e-05,
+      "loss": 1.1094,
+      "step": 3563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.472098584184962,
+      "learning_rate": 4.9250021119129636e-05,
+      "loss": 0.7133,
+      "step": 3564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.6722128245203365,
+      "learning_rate": 4.924900691431005e-05,
+      "loss": 1.2451,
+      "step": 3565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.393707681386393,
+      "learning_rate": 4.924799203465047e-05,
+      "loss": 1.5234,
+      "step": 3566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3476266970178683,
+      "learning_rate": 4.924697648017914e-05,
+      "loss": 0.7939,
+      "step": 3567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5786510508826157,
+      "learning_rate": 4.9245960250924335e-05,
+      "loss": 0.8223,
+      "step": 3568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2660419414048425,
+      "learning_rate": 4.924494334691432e-05,
+      "loss": 0.7656,
+      "step": 3569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4551308446196916,
+      "learning_rate": 4.92439257681774e-05,
+      "loss": 0.7773,
+      "step": 3570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.043420803610319,
+      "learning_rate": 4.92429075147419e-05,
+      "loss": 0.6509,
+      "step": 3571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8373831435016568,
+      "learning_rate": 4.9241888586636155e-05,
+      "loss": 1.2344,
+      "step": 3572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.369656596712534,
+      "learning_rate": 4.924086898388852e-05,
+      "loss": 0.6851,
+      "step": 3573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8271012163376046,
+      "learning_rate": 4.9239848706527366e-05,
+      "loss": 1.4766,
+      "step": 3574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.481523518800769,
+      "learning_rate": 4.923882775458109e-05,
+      "loss": 0.9053,
+      "step": 3575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8519693162827098,
+      "learning_rate": 4.9237806128078105e-05,
+      "loss": 1.0605,
+      "step": 3576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.390969259538036,
+      "learning_rate": 4.923678382704684e-05,
+      "loss": 0.854,
+      "step": 3577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1355776346216246,
+      "learning_rate": 4.9235760851515755e-05,
+      "loss": 1.2344,
+      "step": 3578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.329197750443314,
+      "learning_rate": 4.923473720151331e-05,
+      "loss": 1.0176,
+      "step": 3579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1962504234507625,
+      "learning_rate": 4.923371287706799e-05,
+      "loss": 1.0332,
+      "step": 3580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.863305173761438,
+      "learning_rate": 4.923268787820832e-05,
+      "loss": 1.2188,
+      "step": 3581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.0391874143501605,
+      "learning_rate": 4.923166220496279e-05,
+      "loss": 0.9424,
+      "step": 3582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.522344638397474,
+      "learning_rate": 4.923063585735998e-05,
+      "loss": 0.8188,
+      "step": 3583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3601958002882393,
+      "learning_rate": 4.9229608835428436e-05,
+      "loss": 0.9263,
+      "step": 3584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.408288795135632,
+      "learning_rate": 4.922858113919674e-05,
+      "loss": 0.9492,
+      "step": 3585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7400766683053766,
+      "learning_rate": 4.92275527686935e-05,
+      "loss": 1.4336,
+      "step": 3586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.523415863605724,
+      "learning_rate": 4.922652372394732e-05,
+      "loss": 1.3096,
+      "step": 3587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.376806564165728,
+      "learning_rate": 4.922549400498685e-05,
+      "loss": 1.0889,
+      "step": 3588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5850678762822983,
+      "learning_rate": 4.9224463611840746e-05,
+      "loss": 0.6294,
+      "step": 3589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8684421881132365,
+      "learning_rate": 4.922343254453768e-05,
+      "loss": 1.3105,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.972373522836685,
+      "learning_rate": 4.922240080310635e-05,
+      "loss": 1.4004,
+      "step": 3591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.179427835830176,
+      "learning_rate": 4.922136838757547e-05,
+      "loss": 1.124,
+      "step": 3592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.22066691818571,
+      "learning_rate": 4.9220335297973766e-05,
+      "loss": 1.2012,
+      "step": 3593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.221398692206765,
+      "learning_rate": 4.921930153432999e-05,
+      "loss": 1.0879,
+      "step": 3594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.038823527768989,
+      "learning_rate": 4.921826709667291e-05,
+      "loss": 0.7949,
+      "step": 3595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5871520136156234,
+      "learning_rate": 4.921723198503132e-05,
+      "loss": 1.3486,
+      "step": 3596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7184442808647935,
+      "learning_rate": 4.9216196199434016e-05,
+      "loss": 1.3281,
+      "step": 3597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9835654582364466,
+      "learning_rate": 4.921515973990983e-05,
+      "loss": 0.7778,
+      "step": 3598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.366060410783263,
+      "learning_rate": 4.921412260648761e-05,
+      "loss": 1.8335,
+      "step": 3599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3033626416776647,
+      "learning_rate": 4.921308479919622e-05,
+      "loss": 1.1074,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.433557067352362,
+      "learning_rate": 4.921204631806453e-05,
+      "loss": 1.1201,
+      "step": 3601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.991737398553332,
+      "learning_rate": 4.9211007163121455e-05,
+      "loss": 1.9844,
+      "step": 3602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.101004985216077,
+      "learning_rate": 4.92099673343959e-05,
+      "loss": 1.4434,
+      "step": 3603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.959278296045721,
+      "learning_rate": 4.920892683191681e-05,
+      "loss": 1.2114,
+      "step": 3604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.497543017522991,
+      "learning_rate": 4.920788565571315e-05,
+      "loss": 0.5796,
+      "step": 3605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.137201253182344,
+      "learning_rate": 4.920684380581388e-05,
+      "loss": 1.0918,
+      "step": 3606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.639887788321539,
+      "learning_rate": 4.9205801282248e-05,
+      "loss": 0.4419,
+      "step": 3607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.9746760962889365,
+      "learning_rate": 4.920475808504453e-05,
+      "loss": 1.0371,
+      "step": 3608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.712904385081446,
+      "learning_rate": 4.920371421423249e-05,
+      "loss": 1.7031,
+      "step": 3609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.227809422178445,
+      "learning_rate": 4.9202669669840943e-05,
+      "loss": 1.1621,
+      "step": 3610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4615044803520494,
+      "learning_rate": 4.9201624451898954e-05,
+      "loss": 1.584,
+      "step": 3611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9783811445050747,
+      "learning_rate": 4.92005785604356e-05,
+      "loss": 1.3208,
+      "step": 3612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.594242187129447,
+      "learning_rate": 4.9199531995480005e-05,
+      "loss": 1.1133,
+      "step": 3613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.93492403104291,
+      "learning_rate": 4.919848475706128e-05,
+      "loss": 0.9434,
+      "step": 3614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.5739923253364045,
+      "learning_rate": 4.919743684520858e-05,
+      "loss": 0.7637,
+      "step": 3615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.445840686811616,
+      "learning_rate": 4.919638825995106e-05,
+      "loss": 1.167,
+      "step": 3616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7679535282070695,
+      "learning_rate": 4.919533900131791e-05,
+      "loss": 0.5303,
+      "step": 3617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.256400281854962,
+      "learning_rate": 4.9194289069338314e-05,
+      "loss": 0.7769,
+      "step": 3618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.4899525381742125,
+      "learning_rate": 4.919323846404151e-05,
+      "loss": 0.9941,
+      "step": 3619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.687025338772155,
+      "learning_rate": 4.919218718545673e-05,
+      "loss": 0.9102,
+      "step": 3620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.550700003152919,
+      "learning_rate": 4.9191135233613225e-05,
+      "loss": 0.7517,
+      "step": 3621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.621227096144664,
+      "learning_rate": 4.919008260854028e-05,
+      "loss": 0.8086,
+      "step": 3622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3212180891087786,
+      "learning_rate": 4.9189029310267185e-05,
+      "loss": 0.6328,
+      "step": 3623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.262630614836265,
+      "learning_rate": 4.918797533882325e-05,
+      "loss": 1.1211,
+      "step": 3624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5529491156239335,
+      "learning_rate": 4.91869206942378e-05,
+      "loss": 0.8926,
+      "step": 3625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.931232542492387,
+      "learning_rate": 4.9185865376540206e-05,
+      "loss": 2.5391,
+      "step": 3626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3190458288957023,
+      "learning_rate": 4.918480938575982e-05,
+      "loss": 1.418,
+      "step": 3627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.00067037686879,
+      "learning_rate": 4.9183752721926036e-05,
+      "loss": 1.1436,
+      "step": 3628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4114032406758197,
+      "learning_rate": 4.918269538506826e-05,
+      "loss": 0.7754,
+      "step": 3629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.268307688958755,
+      "learning_rate": 4.918163737521591e-05,
+      "loss": 1.3223,
+      "step": 3630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.219245750042895,
+      "learning_rate": 4.918057869239844e-05,
+      "loss": 1.2539,
+      "step": 3631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.602447634123402,
+      "learning_rate": 4.91795193366453e-05,
+      "loss": 1.5977,
+      "step": 3632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.298586750663352,
+      "learning_rate": 4.917845930798599e-05,
+      "loss": 1.0977,
+      "step": 3633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.221124896741882,
+      "learning_rate": 4.917739860645e-05,
+      "loss": 1.4395,
+      "step": 3634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6568535200683563,
+      "learning_rate": 4.917633723206685e-05,
+      "loss": 1.5391,
+      "step": 3635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0640633154619485,
+      "learning_rate": 4.9175275184866074e-05,
+      "loss": 0.8091,
+      "step": 3636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0150033576908677,
+      "learning_rate": 4.917421246487723e-05,
+      "loss": 0.7417,
+      "step": 3637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.217172059140059,
+      "learning_rate": 4.9173149072129896e-05,
+      "loss": 1.2695,
+      "step": 3638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5602061541818073,
+      "learning_rate": 4.917208500665367e-05,
+      "loss": 0.7451,
+      "step": 3639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.040943025804637,
+      "learning_rate": 4.9171020268478154e-05,
+      "loss": 1.1328,
+      "step": 3640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.75988691512927,
+      "learning_rate": 4.916995485763298e-05,
+      "loss": 0.8066,
+      "step": 3641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.328238977292082,
+      "learning_rate": 4.9168888774147805e-05,
+      "loss": 1.5254,
+      "step": 3642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2670676012363282,
+      "learning_rate": 4.916782201805229e-05,
+      "loss": 0.9023,
+      "step": 3643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6351163388941363,
+      "learning_rate": 4.916675458937614e-05,
+      "loss": 1.082,
+      "step": 3644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6841686709827928,
+      "learning_rate": 4.9165686488149036e-05,
+      "loss": 0.2917,
+      "step": 3645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.213408043512027,
+      "learning_rate": 4.916461771440072e-05,
+      "loss": 0.7083,
+      "step": 3646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1247015350771457,
+      "learning_rate": 4.916354826816093e-05,
+      "loss": 0.8701,
+      "step": 3647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.393673772704225,
+      "learning_rate": 4.9162478149459427e-05,
+      "loss": 1.2178,
+      "step": 3648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.971122355979171,
+      "learning_rate": 4.9161407358326e-05,
+      "loss": 1.2363,
+      "step": 3649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.728195135449847,
+      "learning_rate": 4.9160335894790433e-05,
+      "loss": 1.2852,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8191747510318454,
+      "learning_rate": 4.9159263758882555e-05,
+      "loss": 1.3496,
+      "step": 3651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.85671506472006,
+      "learning_rate": 4.915819095063221e-05,
+      "loss": 1.251,
+      "step": 3652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0970128895753404,
+      "learning_rate": 4.915711747006924e-05,
+      "loss": 1.3574,
+      "step": 3653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.848492534898856,
+      "learning_rate": 4.915604331722352e-05,
+      "loss": 1.0098,
+      "step": 3654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.747198202591749,
+      "learning_rate": 4.9154968492124955e-05,
+      "loss": 1.2812,
+      "step": 3655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9382092808097482,
+      "learning_rate": 4.915389299480345e-05,
+      "loss": 1.1504,
+      "step": 3656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.189309896604146,
+      "learning_rate": 4.915281682528894e-05,
+      "loss": 0.7319,
+      "step": 3657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6541970790715603,
+      "learning_rate": 4.9151739983611364e-05,
+      "loss": 1.291,
+      "step": 3658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6707445396703893,
+      "learning_rate": 4.91506624698007e-05,
+      "loss": 1.4902,
+      "step": 3659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7847419847995822,
+      "learning_rate": 4.914958428388693e-05,
+      "loss": 0.9189,
+      "step": 3660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.933700122931244,
+      "learning_rate": 4.914850542590006e-05,
+      "loss": 1.4688,
+      "step": 3661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8295999429473402,
+      "learning_rate": 4.9147425895870114e-05,
+      "loss": 0.5818,
+      "step": 3662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.122048217743142,
+      "learning_rate": 4.9146345693827134e-05,
+      "loss": 1.2754,
+      "step": 3663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.347363648928576,
+      "learning_rate": 4.9145264819801184e-05,
+      "loss": 1.1719,
+      "step": 3664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.76072811722095,
+      "learning_rate": 4.914418327382234e-05,
+      "loss": 1.4551,
+      "step": 3665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.51224066703491,
+      "learning_rate": 4.914310105592071e-05,
+      "loss": 1.6719,
+      "step": 3666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6144389509374357,
+      "learning_rate": 4.914201816612641e-05,
+      "loss": 1.2129,
+      "step": 3667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0169869149303317,
+      "learning_rate": 4.914093460446957e-05,
+      "loss": 0.7871,
+      "step": 3668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6199642197881197,
+      "learning_rate": 4.913985037098034e-05,
+      "loss": 1.0767,
+      "step": 3669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0711882994698945,
+      "learning_rate": 4.913876546568892e-05,
+      "loss": 1.5781,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.540867278499021,
+      "learning_rate": 4.913767988862547e-05,
+      "loss": 0.8232,
+      "step": 3671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.918915245072965,
+      "learning_rate": 4.9136593639820215e-05,
+      "loss": 0.9724,
+      "step": 3672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6494511701184074,
+      "learning_rate": 4.913550671930339e-05,
+      "loss": 0.6895,
+      "step": 3673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.050696139047957,
+      "learning_rate": 4.9134419127105236e-05,
+      "loss": 1.4277,
+      "step": 3674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6327578960950073,
+      "learning_rate": 4.913333086325602e-05,
+      "loss": 1.5059,
+      "step": 3675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.287719332493104,
+      "learning_rate": 4.9132241927786035e-05,
+      "loss": 0.9961,
+      "step": 3676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5836621557176733,
+      "learning_rate": 4.913115232072558e-05,
+      "loss": 1.3906,
+      "step": 3677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.437858750768036,
+      "learning_rate": 4.913006204210498e-05,
+      "loss": 1.1387,
+      "step": 3678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.189862396238576,
+      "learning_rate": 4.912897109195458e-05,
+      "loss": 1.25,
+      "step": 3679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.607594291562534,
+      "learning_rate": 4.912787947030473e-05,
+      "loss": 1.1348,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.272139416334563,
+      "learning_rate": 4.912678717718582e-05,
+      "loss": 1.4824,
+      "step": 3681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.562762363307635,
+      "learning_rate": 4.912569421262825e-05,
+      "loss": 1.0684,
+      "step": 3682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.335512121105122,
+      "learning_rate": 4.912460057666242e-05,
+      "loss": 1.415,
+      "step": 3683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6399263030755566,
+      "learning_rate": 4.912350626931879e-05,
+      "loss": 1.1006,
+      "step": 3684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.060826981294712,
+      "learning_rate": 4.912241129062779e-05,
+      "loss": 0.7617,
+      "step": 3685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.07437206649148,
+      "learning_rate": 4.9121315640619916e-05,
+      "loss": 1.4883,
+      "step": 3686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.296175335590712,
+      "learning_rate": 4.912021931932563e-05,
+      "loss": 0.9307,
+      "step": 3687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.368643794935497,
+      "learning_rate": 4.911912232677548e-05,
+      "loss": 1.1094,
+      "step": 3688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8108215452893963,
+      "learning_rate": 4.911802466299996e-05,
+      "loss": 0.8218,
+      "step": 3689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5382451546527984,
+      "learning_rate": 4.911692632802964e-05,
+      "loss": 1.2891,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8041269880372397,
+      "learning_rate": 4.9115827321895074e-05,
+      "loss": 1.2715,
+      "step": 3691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7039845593874987,
+      "learning_rate": 4.9114727644626855e-05,
+      "loss": 1.042,
+      "step": 3692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.682426842236692,
+      "learning_rate": 4.9113627296255585e-05,
+      "loss": 1.0381,
+      "step": 3693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.397966192863893,
+      "learning_rate": 4.911252627681188e-05,
+      "loss": 1.4297,
+      "step": 3694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5104663152784394,
+      "learning_rate": 4.911142458632638e-05,
+      "loss": 1.0557,
+      "step": 3695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.420730338069888,
+      "learning_rate": 4.9110322224829754e-05,
+      "loss": 1.1621,
+      "step": 3696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9419085890827281,
+      "learning_rate": 4.910921919235267e-05,
+      "loss": 1.0762,
+      "step": 3697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.605459690793143,
+      "learning_rate": 4.910811548892584e-05,
+      "loss": 2.0195,
+      "step": 3698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1879744968776067,
+      "learning_rate": 4.910701111457997e-05,
+      "loss": 0.8809,
+      "step": 3699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.10718971366602,
+      "learning_rate": 4.9105906069345784e-05,
+      "loss": 0.6484,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6270406148302645,
+      "learning_rate": 4.910480035325405e-05,
+      "loss": 0.3289,
+      "step": 3701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.491502817267552,
+      "learning_rate": 4.910369396633555e-05,
+      "loss": 0.9258,
+      "step": 3702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6478531904638354,
+      "learning_rate": 4.9102586908621046e-05,
+      "loss": 0.8979,
+      "step": 3703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3553246499362364,
+      "learning_rate": 4.910147918014136e-05,
+      "loss": 1.4363,
+      "step": 3704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.018313116932144,
+      "learning_rate": 4.9100370780927324e-05,
+      "loss": 1.3652,
+      "step": 3705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.7984694027375845,
+      "learning_rate": 4.909926171100978e-05,
+      "loss": 1.0703,
+      "step": 3706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.754377522286718,
+      "learning_rate": 4.909815197041959e-05,
+      "loss": 0.5457,
+      "step": 3707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.807610938924184,
+      "learning_rate": 4.9097041559187645e-05,
+      "loss": 1.4883,
+      "step": 3708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.6178723648692,
+      "learning_rate": 4.909593047734484e-05,
+      "loss": 1.4688,
+      "step": 3709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1599492216255554,
+      "learning_rate": 4.90948187249221e-05,
+      "loss": 0.6284,
+      "step": 3710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.683178195591872,
+      "learning_rate": 4.909370630195036e-05,
+      "loss": 0.999,
+      "step": 3711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.797817541957142,
+      "learning_rate": 4.9092593208460594e-05,
+      "loss": 1.7715,
+      "step": 3712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7123112434670107,
+      "learning_rate": 4.9091479444483756e-05,
+      "loss": 1.0352,
+      "step": 3713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9466588512449343,
+      "learning_rate": 4.909036501005086e-05,
+      "loss": 0.917,
+      "step": 3714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3340697587649597,
+      "learning_rate": 4.908924990519291e-05,
+      "loss": 1.0093,
+      "step": 3715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6858922008091977,
+      "learning_rate": 4.908813412994094e-05,
+      "loss": 1.3574,
+      "step": 3716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7475960367789227,
+      "learning_rate": 4.908701768432601e-05,
+      "loss": 1.4531,
+      "step": 3717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.900143072414004,
+      "learning_rate": 4.9085900568379176e-05,
+      "loss": 0.9004,
+      "step": 3718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.636691477475612,
+      "learning_rate": 4.908478278213154e-05,
+      "loss": 0.6846,
+      "step": 3719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2174736972698224,
+      "learning_rate": 4.90836643256142e-05,
+      "loss": 0.7178,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4342016743570922,
+      "learning_rate": 4.908254519885829e-05,
+      "loss": 1.2061,
+      "step": 3721
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.288356702088307,
+      "learning_rate": 4.9081425401894944e-05,
+      "loss": 1.3398,
+      "step": 3722
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0449026130424266,
+      "learning_rate": 4.908030493475534e-05,
+      "loss": 1.0762,
+      "step": 3723
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7665766571141166,
+      "learning_rate": 4.9079183797470654e-05,
+      "loss": 1.1807,
+      "step": 3724
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9442033784781845,
+      "learning_rate": 4.907806199007208e-05,
+      "loss": 1.4219,
+      "step": 3725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6525374906836494,
+      "learning_rate": 4.907693951259084e-05,
+      "loss": 1.1699,
+      "step": 3726
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.040909705774054,
+      "learning_rate": 4.907581636505818e-05,
+      "loss": 1.4102,
+      "step": 3727
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0437364895185905,
+      "learning_rate": 4.907469254750535e-05,
+      "loss": 0.8799,
+      "step": 3728
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8907391222094003,
+      "learning_rate": 4.9073568059963634e-05,
+      "loss": 0.7275,
+      "step": 3729
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.42548706433101,
+      "learning_rate": 4.907244290246431e-05,
+      "loss": 0.9453,
+      "step": 3730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.708102984291597,
+      "learning_rate": 4.90713170750387e-05,
+      "loss": 0.9929,
+      "step": 3731
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5323211852319525,
+      "learning_rate": 4.907019057771815e-05,
+      "loss": 0.998,
+      "step": 3732
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.427222201629938,
+      "learning_rate": 4.906906341053398e-05,
+      "loss": 0.8818,
+      "step": 3733
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.9838312025800535,
+      "learning_rate": 4.906793557351757e-05,
+      "loss": 0.3508,
+      "step": 3734
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2126230238576716,
+      "learning_rate": 4.906680706670033e-05,
+      "loss": 0.9111,
+      "step": 3735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4839964519881215,
+      "learning_rate": 4.9065677890113636e-05,
+      "loss": 1.2109,
+      "step": 3736
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.729937893367145,
+      "learning_rate": 4.906454804378892e-05,
+      "loss": 0.792,
+      "step": 3737
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9661897027045698,
+      "learning_rate": 4.9063417527757635e-05,
+      "loss": 0.9238,
+      "step": 3738
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.534679303321194,
+      "learning_rate": 4.906228634205123e-05,
+      "loss": 1.4707,
+      "step": 3739
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5403655948503294,
+      "learning_rate": 4.90611544867012e-05,
+      "loss": 1.0938,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.565630531177134,
+      "learning_rate": 4.9060021961739035e-05,
+      "loss": 0.959,
+      "step": 3741
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.420897157214012,
+      "learning_rate": 4.905888876719624e-05,
+      "loss": 0.4995,
+      "step": 3742
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.461125345478366,
+      "learning_rate": 4.905775490310438e-05,
+      "loss": 1.3691,
+      "step": 3743
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9262310996047365,
+      "learning_rate": 4.905662036949499e-05,
+      "loss": 1.2324,
+      "step": 3744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.469269309821217,
+      "learning_rate": 4.9055485166399653e-05,
+      "loss": 0.9678,
+      "step": 3745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4631934122330894,
+      "learning_rate": 4.905434929384995e-05,
+      "loss": 0.5618,
+      "step": 3746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0364238869435933,
+      "learning_rate": 4.9053212751877496e-05,
+      "loss": 0.5161,
+      "step": 3747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7822776353244936,
+      "learning_rate": 4.905207554051393e-05,
+      "loss": 1.5332,
+      "step": 3748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.397988640848193,
+      "learning_rate": 4.9050937659790896e-05,
+      "loss": 0.7979,
+      "step": 3749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.223716829505715,
+      "learning_rate": 4.9049799109740054e-05,
+      "loss": 1.3418,
+      "step": 3750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.895087349583558,
+      "learning_rate": 4.904865989039309e-05,
+      "loss": 0.7549,
+      "step": 3751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.581387447777723,
+      "learning_rate": 4.904752000178172e-05,
+      "loss": 1.6172,
+      "step": 3752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2842492338105456,
+      "learning_rate": 4.904637944393765e-05,
+      "loss": 0.7236,
+      "step": 3753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.26449469378098,
+      "learning_rate": 4.904523821689263e-05,
+      "loss": 1.0645,
+      "step": 3754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.501779459418477,
+      "learning_rate": 4.904409632067842e-05,
+      "loss": 1.5713,
+      "step": 3755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1914768581608923,
+      "learning_rate": 4.904295375532681e-05,
+      "loss": 1.0977,
+      "step": 3756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.19414729332895,
+      "learning_rate": 4.904181052086957e-05,
+      "loss": 1.7812,
+      "step": 3757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.270643429356973,
+      "learning_rate": 4.904066661733853e-05,
+      "loss": 1.8857,
+      "step": 3758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0138390487408366,
+      "learning_rate": 4.9039522044765526e-05,
+      "loss": 1.2695,
+      "step": 3759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.437265359216989,
+      "learning_rate": 4.903837680318242e-05,
+      "loss": 1.8281,
+      "step": 3760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.360189275381007,
+      "learning_rate": 4.903723089262107e-05,
+      "loss": 1.2803,
+      "step": 3761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.686117038196475,
+      "learning_rate": 4.9036084313113365e-05,
+      "loss": 1.4297,
+      "step": 3762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.924339583896731,
+      "learning_rate": 4.903493706469122e-05,
+      "loss": 0.6382,
+      "step": 3763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.242818486898604,
+      "learning_rate": 4.903378914738656e-05,
+      "loss": 1.2812,
+      "step": 3764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.189449985406829,
+      "learning_rate": 4.903264056123134e-05,
+      "loss": 1.1328,
+      "step": 3765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.5851585858812784,
+      "learning_rate": 4.90314913062575e-05,
+      "loss": 1.5527,
+      "step": 3766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.3150459182789085,
+      "learning_rate": 4.9030341382497056e-05,
+      "loss": 1.248,
+      "step": 3767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2243001587322864,
+      "learning_rate": 4.902919078998199e-05,
+      "loss": 0.8135,
+      "step": 3768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.5573509110732475,
+      "learning_rate": 4.902803952874433e-05,
+      "loss": 1.2812,
+      "step": 3769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8776155267321486,
+      "learning_rate": 4.902688759881611e-05,
+      "loss": 0.7759,
+      "step": 3770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.553245647023597,
+      "learning_rate": 4.902573500022938e-05,
+      "loss": 1.293,
+      "step": 3771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6312990451328395,
+      "learning_rate": 4.9024581733016234e-05,
+      "loss": 1.2305,
+      "step": 3772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4691014100110795,
+      "learning_rate": 4.902342779720875e-05,
+      "loss": 1.0391,
+      "step": 3773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.505790436669337,
+      "learning_rate": 4.9022273192839064e-05,
+      "loss": 1.5254,
+      "step": 3774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.716411520731909,
+      "learning_rate": 4.9021117919939286e-05,
+      "loss": 1.5898,
+      "step": 3775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1045403353323433,
+      "learning_rate": 4.9019961978541583e-05,
+      "loss": 1.126,
+      "step": 3776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.867762544051531,
+      "learning_rate": 4.901880536867811e-05,
+      "loss": 1.3613,
+      "step": 3777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1532003979153367,
+      "learning_rate": 4.9017648090381066e-05,
+      "loss": 1.4336,
+      "step": 3778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.884771442816691,
+      "learning_rate": 4.9016490143682646e-05,
+      "loss": 1.5977,
+      "step": 3779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.524526303613974,
+      "learning_rate": 4.901533152861509e-05,
+      "loss": 1.2168,
+      "step": 3780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.300658845328963,
+      "learning_rate": 4.901417224521063e-05,
+      "loss": 0.6738,
+      "step": 3781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5391428773335445,
+      "learning_rate": 4.901301229350153e-05,
+      "loss": 0.9863,
+      "step": 3782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.939162735834508,
+      "learning_rate": 4.901185167352008e-05,
+      "loss": 1.0244,
+      "step": 3783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0625409208919963,
+      "learning_rate": 4.901069038529857e-05,
+      "loss": 1.3105,
+      "step": 3784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.585055659006659,
+      "learning_rate": 4.9009528428869324e-05,
+      "loss": 1.2227,
+      "step": 3785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7111875429740784,
+      "learning_rate": 4.900836580426467e-05,
+      "loss": 0.9619,
+      "step": 3786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.717083562757247,
+      "learning_rate": 4.900720251151697e-05,
+      "loss": 1.5,
+      "step": 3787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0862763661797836,
+      "learning_rate": 4.900603855065861e-05,
+      "loss": 1.0283,
+      "step": 3788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5507959245665655,
+      "learning_rate": 4.9004873921721956e-05,
+      "loss": 1.7188,
+      "step": 3789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2691731866451774,
+      "learning_rate": 4.900370862473944e-05,
+      "loss": 1.8008,
+      "step": 3790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.868342746827104,
+      "learning_rate": 4.900254265974348e-05,
+      "loss": 1.3594,
+      "step": 3791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5334232561146988,
+      "learning_rate": 4.9001376026766535e-05,
+      "loss": 1.3896,
+      "step": 3792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.966911430714492,
+      "learning_rate": 4.900020872584106e-05,
+      "loss": 0.855,
+      "step": 3793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8783241625538305,
+      "learning_rate": 4.899904075699955e-05,
+      "loss": 1.3047,
+      "step": 3794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8186073960731464,
+      "learning_rate": 4.89978721202745e-05,
+      "loss": 1.4434,
+      "step": 3795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.094442294456806,
+      "learning_rate": 4.899670281569845e-05,
+      "loss": 1.2314,
+      "step": 3796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.060897240183133,
+      "learning_rate": 4.899553284330392e-05,
+      "loss": 1.1406,
+      "step": 3797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.3422908463904095,
+      "learning_rate": 4.8994362203123484e-05,
+      "loss": 1.7324,
+      "step": 3798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3680693496124494,
+      "learning_rate": 4.8993190895189714e-05,
+      "loss": 1.9766,
+      "step": 3799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.115883513620939,
+      "learning_rate": 4.899201891953521e-05,
+      "loss": 1.1465,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.249253700840665,
+      "learning_rate": 4.899084627619259e-05,
+      "loss": 0.9648,
+      "step": 3801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2134279708086297,
+      "learning_rate": 4.898967296519448e-05,
+      "loss": 0.9385,
+      "step": 3802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.037762925671545,
+      "learning_rate": 4.8988498986573536e-05,
+      "loss": 0.8994,
+      "step": 3803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3638231397072476,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 1.0908,
+      "step": 3804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.253016780629929,
+      "learning_rate": 4.898614902659386e-05,
+      "loss": 0.5251,
+      "step": 3805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.956182089368886,
+      "learning_rate": 4.8984973045300534e-05,
+      "loss": 0.9102,
+      "step": 3806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3545545635402014,
+      "learning_rate": 4.8983796396515156e-05,
+      "loss": 1.1895,
+      "step": 3807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.094518123820979,
+      "learning_rate": 4.89826190802705e-05,
+      "loss": 1.1094,
+      "step": 3808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.658098711487019,
+      "learning_rate": 4.8981441096599314e-05,
+      "loss": 1.0605,
+      "step": 3809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0110766645418603,
+      "learning_rate": 4.8980262445534375e-05,
+      "loss": 1.2266,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.357845796080758,
+      "learning_rate": 4.89790831271085e-05,
+      "loss": 1.1875,
+      "step": 3811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.725142128560098,
+      "learning_rate": 4.897790314135452e-05,
+      "loss": 1.1792,
+      "step": 3812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.850922258565617,
+      "learning_rate": 4.897672248830524e-05,
+      "loss": 1.5273,
+      "step": 3813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3074878296371013,
+      "learning_rate": 4.897554116799354e-05,
+      "loss": 1.1055,
+      "step": 3814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9968253211586173,
+      "learning_rate": 4.897435918045229e-05,
+      "loss": 0.6853,
+      "step": 3815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.94555944384386,
+      "learning_rate": 4.8973176525714375e-05,
+      "loss": 1.1294,
+      "step": 3816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.794875298517831,
+      "learning_rate": 4.897199320381273e-05,
+      "loss": 0.7163,
+      "step": 3817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.40754110204167,
+      "learning_rate": 4.897080921478027e-05,
+      "loss": 0.9956,
+      "step": 3818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.953374261718951,
+      "learning_rate": 4.896962455864994e-05,
+      "loss": 1.0144,
+      "step": 3819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.410428277842657,
+      "learning_rate": 4.896843923545472e-05,
+      "loss": 1.0645,
+      "step": 3820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.290982732008658,
+      "learning_rate": 4.89672532452276e-05,
+      "loss": 0.8516,
+      "step": 3821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.607879301373245,
+      "learning_rate": 4.896606658800157e-05,
+      "loss": 0.7014,
+      "step": 3822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0743613903476157,
+      "learning_rate": 4.896487926380967e-05,
+      "loss": 0.7424,
+      "step": 3823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9913980822129123,
+      "learning_rate": 4.896369127268493e-05,
+      "loss": 0.9336,
+      "step": 3824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.730998717736216,
+      "learning_rate": 4.896250261466042e-05,
+      "loss": 1.1836,
+      "step": 3825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9412960117954645,
+      "learning_rate": 4.8961313289769216e-05,
+      "loss": 1.0029,
+      "step": 3826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.169088944342742,
+      "learning_rate": 4.896012329804443e-05,
+      "loss": 1.1895,
+      "step": 3827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9021579277344998,
+      "learning_rate": 4.895893263951915e-05,
+      "loss": 1.3418,
+      "step": 3828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.514982923642265,
+      "learning_rate": 4.895774131422654e-05,
+      "loss": 1.3828,
+      "step": 3829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1956691403010513,
+      "learning_rate": 4.895654932219974e-05,
+      "loss": 1.166,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4424691090690316,
+      "learning_rate": 4.895535666347192e-05,
+      "loss": 0.5908,
+      "step": 3831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.403141107862619,
+      "learning_rate": 4.895416333807628e-05,
+      "loss": 0.9321,
+      "step": 3832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.561209711163398,
+      "learning_rate": 4.8952969346046026e-05,
+      "loss": 1.0723,
+      "step": 3833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.254393209283915,
+      "learning_rate": 4.895177468741439e-05,
+      "loss": 1.1943,
+      "step": 3834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.566474902020079,
+      "learning_rate": 4.895057936221461e-05,
+      "loss": 0.9141,
+      "step": 3835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.890113909103372,
+      "learning_rate": 4.8949383370479956e-05,
+      "loss": 1.623,
+      "step": 3836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.571354887065988,
+      "learning_rate": 4.8948186712243715e-05,
+      "loss": 1.1016,
+      "step": 3837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3342628217142565,
+      "learning_rate": 4.894698938753919e-05,
+      "loss": 1.0693,
+      "step": 3838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6361370886203925,
+      "learning_rate": 4.8945791396399686e-05,
+      "loss": 1.541,
+      "step": 3839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0712069307877563,
+      "learning_rate": 4.894459273885858e-05,
+      "loss": 0.8662,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.182255916864068,
+      "learning_rate": 4.894339341494918e-05,
+      "loss": 0.4646,
+      "step": 3841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.839839845979754,
+      "learning_rate": 4.89421934247049e-05,
+      "loss": 1.2822,
+      "step": 3842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7277268161748163,
+      "learning_rate": 4.8940992768159126e-05,
+      "loss": 1.2969,
+      "step": 3843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.744436142942299,
+      "learning_rate": 4.893979144534527e-05,
+      "loss": 1.3652,
+      "step": 3844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6144523334487726,
+      "learning_rate": 4.893858945629676e-05,
+      "loss": 1.2617,
+      "step": 3845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6503274217049102,
+      "learning_rate": 4.893738680104705e-05,
+      "loss": 0.9458,
+      "step": 3846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4818248976388895,
+      "learning_rate": 4.8936183479629616e-05,
+      "loss": 0.9409,
+      "step": 3847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4165475640599365,
+      "learning_rate": 4.893497949207794e-05,
+      "loss": 1.1748,
+      "step": 3848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.0063040093450555,
+      "learning_rate": 4.893377483842552e-05,
+      "loss": 1.7227,
+      "step": 3849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1333189422400176,
+      "learning_rate": 4.8932569518705896e-05,
+      "loss": 1.3711,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.597239119114871,
+      "learning_rate": 4.89313635329526e-05,
+      "loss": 0.9844,
+      "step": 3851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.031487982770705,
+      "learning_rate": 4.893015688119921e-05,
+      "loss": 0.708,
+      "step": 3852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.191066138134676,
+      "learning_rate": 4.892894956347929e-05,
+      "loss": 2.2461,
+      "step": 3853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0021354480620124,
+      "learning_rate": 4.8927741579826444e-05,
+      "loss": 0.8867,
+      "step": 3854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.289008246184641,
+      "learning_rate": 4.892653293027429e-05,
+      "loss": 1.4766,
+      "step": 3855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2554545514063173,
+      "learning_rate": 4.8925323614856466e-05,
+      "loss": 1.1133,
+      "step": 3856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.12826210968443,
+      "learning_rate": 4.892411363360663e-05,
+      "loss": 1.6953,
+      "step": 3857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.191293085323462,
+      "learning_rate": 4.892290298655844e-05,
+      "loss": 0.8052,
+      "step": 3858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1891865359555327,
+      "learning_rate": 4.892169167374561e-05,
+      "loss": 0.8926,
+      "step": 3859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.782962866247994,
+      "learning_rate": 4.892047969520184e-05,
+      "loss": 1.0098,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6128958516261824,
+      "learning_rate": 4.891926705096085e-05,
+      "loss": 1.5498,
+      "step": 3861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.981877520345915,
+      "learning_rate": 4.89180537410564e-05,
+      "loss": 0.7388,
+      "step": 3862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3356142981880947,
+      "learning_rate": 4.891683976552225e-05,
+      "loss": 1.2832,
+      "step": 3863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1861755721355385,
+      "learning_rate": 4.891562512439219e-05,
+      "loss": 1.3066,
+      "step": 3864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.330687987462836,
+      "learning_rate": 4.891440981770001e-05,
+      "loss": 1.1543,
+      "step": 3865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.324150225486701,
+      "learning_rate": 4.891319384547955e-05,
+      "loss": 1.1895,
+      "step": 3866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.203820380223752,
+      "learning_rate": 4.891197720776464e-05,
+      "loss": 1.1035,
+      "step": 3867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.512163959325512,
+      "learning_rate": 4.8910759904589135e-05,
+      "loss": 1.1055,
+      "step": 3868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4227583628984117,
+      "learning_rate": 4.890954193598691e-05,
+      "loss": 1.3926,
+      "step": 3869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8393414878337,
+      "learning_rate": 4.8908323301991874e-05,
+      "loss": 1.0498,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0578771460575163,
+      "learning_rate": 4.8907104002637934e-05,
+      "loss": 0.9531,
+      "step": 3871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2866833922920926,
+      "learning_rate": 4.8905884037959014e-05,
+      "loss": 1.0422,
+      "step": 3872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.395876698977515,
+      "learning_rate": 4.8904663407989084e-05,
+      "loss": 1.0039,
+      "step": 3873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4505496751396714,
+      "learning_rate": 4.89034421127621e-05,
+      "loss": 0.7832,
+      "step": 3874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6000015112065546,
+      "learning_rate": 4.890222015231205e-05,
+      "loss": 0.5903,
+      "step": 3875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5484482557166497,
+      "learning_rate": 4.890099752667294e-05,
+      "loss": 0.5603,
+      "step": 3876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.699988888466278,
+      "learning_rate": 4.889977423587881e-05,
+      "loss": 1.6777,
+      "step": 3877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.575239458934283,
+      "learning_rate": 4.8898550279963684e-05,
+      "loss": 1.1514,
+      "step": 3878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.954281644821067,
+      "learning_rate": 4.889732565896164e-05,
+      "loss": 1.3242,
+      "step": 3879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3812926401240433,
+      "learning_rate": 4.889610037290675e-05,
+      "loss": 0.8203,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5608303775383767,
+      "learning_rate": 4.889487442183311e-05,
+      "loss": 0.4214,
+      "step": 3881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2689775589244667,
+      "learning_rate": 4.8893647805774844e-05,
+      "loss": 0.436,
+      "step": 3882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3942365266007526,
+      "learning_rate": 4.8892420524766084e-05,
+      "loss": 0.9414,
+      "step": 3883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.103743607199155,
+      "learning_rate": 4.889119257884099e-05,
+      "loss": 1.2959,
+      "step": 3884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3487431553796227,
+      "learning_rate": 4.888996396803374e-05,
+      "loss": 0.8774,
+      "step": 3885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.820344550769661,
+      "learning_rate": 4.8888734692378505e-05,
+      "loss": 1.2891,
+      "step": 3886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9726273227114834,
+      "learning_rate": 4.888750475190952e-05,
+      "loss": 0.8889,
+      "step": 3887
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6971944679415105,
+      "learning_rate": 4.888627414666099e-05,
+      "loss": 1.5117,
+      "step": 3888
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.513777060038233,
+      "learning_rate": 4.8885042876667186e-05,
+      "loss": 0.7354,
+      "step": 3889
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.862793273747858,
+      "learning_rate": 4.888381094196236e-05,
+      "loss": 1.2349,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3085880012845985,
+      "learning_rate": 4.88825783425808e-05,
+      "loss": 0.9443,
+      "step": 3891
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.276370350527693,
+      "learning_rate": 4.88813450785568e-05,
+      "loss": 0.8623,
+      "step": 3892
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.03833048401744,
+      "learning_rate": 4.888011114992469e-05,
+      "loss": 1.0283,
+      "step": 3893
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.363368676384645,
+      "learning_rate": 4.887887655671881e-05,
+      "loss": 0.8513,
+      "step": 3894
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.137432848451109,
+      "learning_rate": 4.887764129897352e-05,
+      "loss": 1.1445,
+      "step": 3895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7979004754123227,
+      "learning_rate": 4.8876405376723194e-05,
+      "loss": 1.2344,
+      "step": 3896
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.170485017427318,
+      "learning_rate": 4.887516879000222e-05,
+      "loss": 1.0006,
+      "step": 3897
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.318258890805699,
+      "learning_rate": 4.887393153884502e-05,
+      "loss": 1.3984,
+      "step": 3898
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.527902343903922,
+      "learning_rate": 4.887269362328603e-05,
+      "loss": 1.2588,
+      "step": 3899
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.097086471348993,
+      "learning_rate": 4.8871455043359696e-05,
+      "loss": 0.9355,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.473165824547139,
+      "learning_rate": 4.887021579910048e-05,
+      "loss": 1.1855,
+      "step": 3901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0965709755817605,
+      "learning_rate": 4.8868975890542876e-05,
+      "loss": 0.4443,
+      "step": 3902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8201255842071045,
+      "learning_rate": 4.886773531772138e-05,
+      "loss": 0.8164,
+      "step": 3903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8314900479647873,
+      "learning_rate": 4.8866494080670544e-05,
+      "loss": 1.0132,
+      "step": 3904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.935950303758738,
+      "learning_rate": 4.886525217942489e-05,
+      "loss": 2.2559,
+      "step": 3905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9860574393441564,
+      "learning_rate": 4.886400961401898e-05,
+      "loss": 0.9895,
+      "step": 3906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.320368396843551,
+      "learning_rate": 4.88627663844874e-05,
+      "loss": 1.2324,
+      "step": 3907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.776260236355628,
+      "learning_rate": 4.886152249086474e-05,
+      "loss": 0.7057,
+      "step": 3908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.440829717082736,
+      "learning_rate": 4.886027793318563e-05,
+      "loss": 0.4976,
+      "step": 3909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.973552949639834,
+      "learning_rate": 4.88590327114847e-05,
+      "loss": 1.8047,
+      "step": 3910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4256846502379603,
+      "learning_rate": 4.88577868257966e-05,
+      "loss": 1.502,
+      "step": 3911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.057307765903142,
+      "learning_rate": 4.8856540276156e-05,
+      "loss": 1.2207,
+      "step": 3912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2333207210735027,
+      "learning_rate": 4.88552930625976e-05,
+      "loss": 1.1704,
+      "step": 3913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.284032381390306,
+      "learning_rate": 4.88540451851561e-05,
+      "loss": 2.373,
+      "step": 3914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.732543593579837,
+      "learning_rate": 4.885279664386625e-05,
+      "loss": 1.1836,
+      "step": 3915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2355548437904016,
+      "learning_rate": 4.885154743876277e-05,
+      "loss": 1.014,
+      "step": 3916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4399513242900532,
+      "learning_rate": 4.8850297569880434e-05,
+      "loss": 1.6191,
+      "step": 3917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.353527332931881,
+      "learning_rate": 4.884904703725403e-05,
+      "loss": 1.5742,
+      "step": 3918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.758958900782391,
+      "learning_rate": 4.884779584091834e-05,
+      "loss": 0.9941,
+      "step": 3919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6957282382912995,
+      "learning_rate": 4.884654398090822e-05,
+      "loss": 1.1377,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.694082089395892,
+      "learning_rate": 4.884529145725848e-05,
+      "loss": 0.9941,
+      "step": 3921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.880023161632189,
+      "learning_rate": 4.884403827000398e-05,
+      "loss": 1.0137,
+      "step": 3922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.765369554349878,
+      "learning_rate": 4.884278441917961e-05,
+      "loss": 1.498,
+      "step": 3923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.855656697805023,
+      "learning_rate": 4.884152990482026e-05,
+      "loss": 1.4316,
+      "step": 3924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.8399424651999885,
+      "learning_rate": 4.884027472696083e-05,
+      "loss": 1.2441,
+      "step": 3925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.354532594529669,
+      "learning_rate": 4.8839018885636264e-05,
+      "loss": 0.3286,
+      "step": 3926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.891202412855639,
+      "learning_rate": 4.883776238088151e-05,
+      "loss": 0.7112,
+      "step": 3927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7370189412456312,
+      "learning_rate": 4.883650521273153e-05,
+      "loss": 0.8018,
+      "step": 3928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2739483760721164,
+      "learning_rate": 4.883524738122132e-05,
+      "loss": 0.6855,
+      "step": 3929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.399049109966768,
+      "learning_rate": 4.883398888638586e-05,
+      "loss": 1.2578,
+      "step": 3930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.807838306510076,
+      "learning_rate": 4.883272972826021e-05,
+      "loss": 1.3242,
+      "step": 3931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4429018856275695,
+      "learning_rate": 4.883146990687939e-05,
+      "loss": 0.5195,
+      "step": 3932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4938804114333073,
+      "learning_rate": 4.883020942227846e-05,
+      "loss": 1.208,
+      "step": 3933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9525148425948533,
+      "learning_rate": 4.8828948274492506e-05,
+      "loss": 1.334,
+      "step": 3934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5883122629690543,
+      "learning_rate": 4.8827686463556624e-05,
+      "loss": 1.0986,
+      "step": 3935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0776246120938393,
+      "learning_rate": 4.882642398950593e-05,
+      "loss": 0.8545,
+      "step": 3936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.992282318126836,
+      "learning_rate": 4.882516085237555e-05,
+      "loss": 1.2754,
+      "step": 3937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.344366740725696,
+      "learning_rate": 4.882389705220064e-05,
+      "loss": 1.0967,
+      "step": 3938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2620586871115362,
+      "learning_rate": 4.882263258901638e-05,
+      "loss": 1.4941,
+      "step": 3939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.094934905833504,
+      "learning_rate": 4.882136746285795e-05,
+      "loss": 1.21,
+      "step": 3940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.218166005821549,
+      "learning_rate": 4.882010167376057e-05,
+      "loss": 1.0977,
+      "step": 3941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.716223363283675,
+      "learning_rate": 4.881883522175945e-05,
+      "loss": 1.4941,
+      "step": 3942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1453276720503767,
+      "learning_rate": 4.881756810688984e-05,
+      "loss": 0.575,
+      "step": 3943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5808332610220113,
+      "learning_rate": 4.881630032918701e-05,
+      "loss": 0.8481,
+      "step": 3944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.279421347755396,
+      "learning_rate": 4.881503188868624e-05,
+      "loss": 1.21,
+      "step": 3945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.145363335310774,
+      "learning_rate": 4.8813762785422826e-05,
+      "loss": 1.3711,
+      "step": 3946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4102203509000124,
+      "learning_rate": 4.881249301943208e-05,
+      "loss": 0.9678,
+      "step": 3947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4551811591160373,
+      "learning_rate": 4.881122259074935e-05,
+      "loss": 1.3555,
+      "step": 3948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9452002312922276,
+      "learning_rate": 4.880995149941e-05,
+      "loss": 0.8887,
+      "step": 3949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7532909815242688,
+      "learning_rate": 4.880867974544938e-05,
+      "loss": 1.2598,
+      "step": 3950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.108803436775954,
+      "learning_rate": 4.8807407328902896e-05,
+      "loss": 1.2109,
+      "step": 3951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.981379738469841,
+      "learning_rate": 4.880613424980597e-05,
+      "loss": 0.6455,
+      "step": 3952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3767925729528474,
+      "learning_rate": 4.880486050819401e-05,
+      "loss": 0.8721,
+      "step": 3953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.206817508096675,
+      "learning_rate": 4.880358610410247e-05,
+      "loss": 0.5037,
+      "step": 3954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.703848444151541,
+      "learning_rate": 4.880231103756683e-05,
+      "loss": 0.9102,
+      "step": 3955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2459998823014273,
+      "learning_rate": 4.880103530862256e-05,
+      "loss": 1.5645,
+      "step": 3956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.215867168178994,
+      "learning_rate": 4.8799758917305164e-05,
+      "loss": 1.4336,
+      "step": 3957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.713107513040966,
+      "learning_rate": 4.879848186365017e-05,
+      "loss": 1.0938,
+      "step": 3958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.470389341857981,
+      "learning_rate": 4.87972041476931e-05,
+      "loss": 1.2451,
+      "step": 3959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8716653416675058,
+      "learning_rate": 4.8795925769469544e-05,
+      "loss": 0.4971,
+      "step": 3960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3713895559794675,
+      "learning_rate": 4.8794646729015055e-05,
+      "loss": 1.209,
+      "step": 3961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.5054019717312555,
+      "learning_rate": 4.879336702636523e-05,
+      "loss": 1.1592,
+      "step": 3962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.175621478859524,
+      "learning_rate": 4.8792086661555695e-05,
+      "loss": 0.9458,
+      "step": 3963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.412850997150984,
+      "learning_rate": 4.879080563462207e-05,
+      "loss": 1.2676,
+      "step": 3964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.467788001325711,
+      "learning_rate": 4.8789523945600004e-05,
+      "loss": 1.3472,
+      "step": 3965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1014587231241166,
+      "learning_rate": 4.8788241594525175e-05,
+      "loss": 0.8789,
+      "step": 3966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.752150801542454,
+      "learning_rate": 4.8786958581433275e-05,
+      "loss": 1.6045,
+      "step": 3967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.070673719905163,
+      "learning_rate": 4.8785674906359994e-05,
+      "loss": 1.0122,
+      "step": 3968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2891265790410227,
+      "learning_rate": 4.878439056934106e-05,
+      "loss": 0.6519,
+      "step": 3969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.710220003922488,
+      "learning_rate": 4.8783105570412225e-05,
+      "loss": 1.3086,
+      "step": 3970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.943221377382486,
+      "learning_rate": 4.878181990960924e-05,
+      "loss": 0.8486,
+      "step": 3971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.18102155215998,
+      "learning_rate": 4.8780533586967895e-05,
+      "loss": 1.168,
+      "step": 3972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.550493092070512,
+      "learning_rate": 4.877924660252397e-05,
+      "loss": 1.9629,
+      "step": 3973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8017053928304687,
+      "learning_rate": 4.87779589563133e-05,
+      "loss": 1.3516,
+      "step": 3974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.912348559037572,
+      "learning_rate": 4.877667064837171e-05,
+      "loss": 1.25,
+      "step": 3975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.977798485644379,
+      "learning_rate": 4.8775381678735065e-05,
+      "loss": 1.2881,
+      "step": 3976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7453377033351645,
+      "learning_rate": 4.8774092047439214e-05,
+      "loss": 0.5835,
+      "step": 3977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.716281515641705,
+      "learning_rate": 4.877280175452007e-05,
+      "loss": 0.8887,
+      "step": 3978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.123735724946219,
+      "learning_rate": 4.877151080001352e-05,
+      "loss": 1.207,
+      "step": 3979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.970889978848708,
+      "learning_rate": 4.877021918395551e-05,
+      "loss": 0.7109,
+      "step": 3980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4216317331516533,
+      "learning_rate": 4.8768926906381975e-05,
+      "loss": 1.3848,
+      "step": 3981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.901053549133767,
+      "learning_rate": 4.876763396732888e-05,
+      "loss": 1.3281,
+      "step": 3982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.129107193401452,
+      "learning_rate": 4.8766340366832206e-05,
+      "loss": 1.2598,
+      "step": 3983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9295218083833878,
+      "learning_rate": 4.876504610492796e-05,
+      "loss": 1.4766,
+      "step": 3984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.708395609995504,
+      "learning_rate": 4.876375118165215e-05,
+      "loss": 1.4238,
+      "step": 3985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.284175025140884,
+      "learning_rate": 4.8762455597040825e-05,
+      "loss": 1.5879,
+      "step": 3986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.696724119320074,
+      "learning_rate": 4.8761159351130025e-05,
+      "loss": 1.3555,
+      "step": 3987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.300236005724658,
+      "learning_rate": 4.875986244395584e-05,
+      "loss": 1.0884,
+      "step": 3988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9679237327257673,
+      "learning_rate": 4.8758564875554355e-05,
+      "loss": 1.3789,
+      "step": 3989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.42484778045909,
+      "learning_rate": 4.875726664596168e-05,
+      "loss": 1.5439,
+      "step": 3990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.892041751876795,
+      "learning_rate": 4.875596775521395e-05,
+      "loss": 1.3438,
+      "step": 3991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6715383931106547,
+      "learning_rate": 4.87546682033473e-05,
+      "loss": 1.0527,
+      "step": 3992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.597945600429266,
+      "learning_rate": 4.875336799039791e-05,
+      "loss": 1.3086,
+      "step": 3993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.94646433358556,
+      "learning_rate": 4.8752067116401965e-05,
+      "loss": 1.167,
+      "step": 3994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.206961813394541,
+      "learning_rate": 4.8750765581395655e-05,
+      "loss": 1.4023,
+      "step": 3995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.782334191385224,
+      "learning_rate": 4.874946338541521e-05,
+      "loss": 0.7095,
+      "step": 3996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7782271317312495,
+      "learning_rate": 4.874816052849686e-05,
+      "loss": 0.9951,
+      "step": 3997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.932444821497171,
+      "learning_rate": 4.874685701067688e-05,
+      "loss": 1.1309,
+      "step": 3998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6481212325853742,
+      "learning_rate": 4.874555283199153e-05,
+      "loss": 1.4258,
+      "step": 3999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9927866984352747,
+      "learning_rate": 4.874424799247711e-05,
+      "loss": 0.51,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.872776404693398,
+      "learning_rate": 4.874294249216994e-05,
+      "loss": 1.3223,
+      "step": 4001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.657275282508847,
+      "learning_rate": 4.874163633110635e-05,
+      "loss": 1.5215,
+      "step": 4002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6580641890296235,
+      "learning_rate": 4.874032950932268e-05,
+      "loss": 1.0859,
+      "step": 4003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6291944216862408,
+      "learning_rate": 4.873902202685531e-05,
+      "loss": 1.0215,
+      "step": 4004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.212440123542837,
+      "learning_rate": 4.873771388374061e-05,
+      "loss": 1.332,
+      "step": 4005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.710669059127285,
+      "learning_rate": 4.8736405080015005e-05,
+      "loss": 1.3887,
+      "step": 4006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.219905892719434,
+      "learning_rate": 4.873509561571491e-05,
+      "loss": 1.2812,
+      "step": 4007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2795538014952363,
+      "learning_rate": 4.873378549087676e-05,
+      "loss": 1.124,
+      "step": 4008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4713027889805406,
+      "learning_rate": 4.873247470553703e-05,
+      "loss": 0.7915,
+      "step": 4009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5128022793520275,
+      "learning_rate": 4.873116325973218e-05,
+      "loss": 1.0947,
+      "step": 4010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7904543355634304,
+      "learning_rate": 4.872985115349873e-05,
+      "loss": 0.9062,
+      "step": 4011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.88678588333183,
+      "learning_rate": 4.872853838687318e-05,
+      "loss": 1.1841,
+      "step": 4012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7315545703793145,
+      "learning_rate": 4.872722495989206e-05,
+      "loss": 1.0605,
+      "step": 4013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.272690981576682,
+      "learning_rate": 4.872591087259193e-05,
+      "loss": 1.0977,
+      "step": 4014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6959855813457165,
+      "learning_rate": 4.8724596125009364e-05,
+      "loss": 0.553,
+      "step": 4015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.70742278046296,
+      "learning_rate": 4.872328071718094e-05,
+      "loss": 1.2852,
+      "step": 4016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.052954113030771,
+      "learning_rate": 4.872196464914328e-05,
+      "loss": 1.0781,
+      "step": 4017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.44890508514067,
+      "learning_rate": 4.872064792093299e-05,
+      "loss": 1.0322,
+      "step": 4018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.063994952218914,
+      "learning_rate": 4.871933053258674e-05,
+      "loss": 1.0225,
+      "step": 4019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.659969284052365,
+      "learning_rate": 4.871801248414116e-05,
+      "loss": 1.1309,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8381412181626513,
+      "learning_rate": 4.871669377563296e-05,
+      "loss": 0.8032,
+      "step": 4021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.248848465694843,
+      "learning_rate": 4.871537440709883e-05,
+      "loss": 1.168,
+      "step": 4022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.130490409209363,
+      "learning_rate": 4.871405437857547e-05,
+      "loss": 1.1597,
+      "step": 4023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.106248021234183,
+      "learning_rate": 4.8712733690099645e-05,
+      "loss": 1.6856,
+      "step": 4024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9462129951648808,
+      "learning_rate": 4.871141234170808e-05,
+      "loss": 1.6113,
+      "step": 4025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.127040687106723,
+      "learning_rate": 4.8710090333437575e-05,
+      "loss": 1.5566,
+      "step": 4026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.330141666182286,
+      "learning_rate": 4.8708767665324905e-05,
+      "loss": 0.896,
+      "step": 4027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.873378123972456,
+      "learning_rate": 4.870744433740688e-05,
+      "loss": 1.0957,
+      "step": 4028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0219546057493,
+      "learning_rate": 4.870612034972034e-05,
+      "loss": 1.4727,
+      "step": 4029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.663066689086332,
+      "learning_rate": 4.8704795702302116e-05,
+      "loss": 1.2759,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.654600798718783,
+      "learning_rate": 4.870347039518908e-05,
+      "loss": 0.9375,
+      "step": 4031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.313957181747165,
+      "learning_rate": 4.870214442841811e-05,
+      "loss": 1.375,
+      "step": 4032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.798263013471248,
+      "learning_rate": 4.87008178020261e-05,
+      "loss": 1.1426,
+      "step": 4033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6564784232264387,
+      "learning_rate": 4.869949051604999e-05,
+      "loss": 0.7959,
+      "step": 4034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9048012402991463,
+      "learning_rate": 4.869816257052671e-05,
+      "loss": 1.0928,
+      "step": 4035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8589116486307997,
+      "learning_rate": 4.86968339654932e-05,
+      "loss": 0.6519,
+      "step": 4036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2554546571138667,
+      "learning_rate": 4.869550470098646e-05,
+      "loss": 1.2188,
+      "step": 4037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5559008222271697,
+      "learning_rate": 4.869417477704347e-05,
+      "loss": 1.0117,
+      "step": 4038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.753680502508909,
+      "learning_rate": 4.869284419370123e-05,
+      "loss": 1.0752,
+      "step": 4039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6887972861206157,
+      "learning_rate": 4.8691512950996794e-05,
+      "loss": 1.167,
+      "step": 4040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.566396166745991,
+      "learning_rate": 4.869018104896719e-05,
+      "loss": 0.9673,
+      "step": 4041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.263115956223932,
+      "learning_rate": 4.86888484876495e-05,
+      "loss": 0.9399,
+      "step": 4042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.060565735066235,
+      "learning_rate": 4.868751526708079e-05,
+      "loss": 0.54,
+      "step": 4043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.507552173359697,
+      "learning_rate": 4.868618138729818e-05,
+      "loss": 1.0742,
+      "step": 4044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4242629553366175,
+      "learning_rate": 4.8684846848338785e-05,
+      "loss": 0.8804,
+      "step": 4045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9844322496701376,
+      "learning_rate": 4.868351165023974e-05,
+      "loss": 1.2314,
+      "step": 4046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3987945390296592,
+      "learning_rate": 4.8682175793038206e-05,
+      "loss": 1.084,
+      "step": 4047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8145464763209374,
+      "learning_rate": 4.8680839276771366e-05,
+      "loss": 1.2871,
+      "step": 4048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8359454215670195,
+      "learning_rate": 4.867950210147641e-05,
+      "loss": 1.0996,
+      "step": 4049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1097262619858284,
+      "learning_rate": 4.867816426719055e-05,
+      "loss": 0.9668,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5299988468454018,
+      "learning_rate": 4.8676825773951015e-05,
+      "loss": 1.2578,
+      "step": 4051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.309604170719771,
+      "learning_rate": 4.8675486621795055e-05,
+      "loss": 0.8208,
+      "step": 4052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7562013848280005,
+      "learning_rate": 4.867414681075995e-05,
+      "loss": 1.0205,
+      "step": 4053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3141904271405256,
+      "learning_rate": 4.867280634088297e-05,
+      "loss": 1.1191,
+      "step": 4054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.193131418857438,
+      "learning_rate": 4.8671465212201426e-05,
+      "loss": 0.8477,
+      "step": 4055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.926071863430902,
+      "learning_rate": 4.867012342475264e-05,
+      "loss": 1.3848,
+      "step": 4056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5652040625936925,
+      "learning_rate": 4.866878097857396e-05,
+      "loss": 0.7053,
+      "step": 4057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.919723723898367,
+      "learning_rate": 4.8667437873702744e-05,
+      "loss": 1.0278,
+      "step": 4058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.034779504284386,
+      "learning_rate": 4.866609411017636e-05,
+      "loss": 1.4277,
+      "step": 4059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.134213980566274,
+      "learning_rate": 4.866474968803222e-05,
+      "loss": 1.5547,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.614118000177834,
+      "learning_rate": 4.866340460730772e-05,
+      "loss": 1.6182,
+      "step": 4061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7328489369673483,
+      "learning_rate": 4.866205886804031e-05,
+      "loss": 0.7588,
+      "step": 4062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0126968952589035,
+      "learning_rate": 4.866071247026743e-05,
+      "loss": 1.4546,
+      "step": 4063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9477275885819145,
+      "learning_rate": 4.865936541402656e-05,
+      "loss": 0.5264,
+      "step": 4064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.061549915156636,
+      "learning_rate": 4.865801769935517e-05,
+      "loss": 1.1162,
+      "step": 4065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.837441369360609,
+      "learning_rate": 4.865666932629079e-05,
+      "loss": 1.8887,
+      "step": 4066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.440993617473355,
+      "learning_rate": 4.865532029487093e-05,
+      "loss": 0.9834,
+      "step": 4067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.548782843813691,
+      "learning_rate": 4.8653970605133134e-05,
+      "loss": 1.0479,
+      "step": 4068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.184339554788261,
+      "learning_rate": 4.865262025711497e-05,
+      "loss": 0.9678,
+      "step": 4069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.952976593834061,
+      "learning_rate": 4.865126925085401e-05,
+      "loss": 1.4336,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8603418935480143,
+      "learning_rate": 4.864991758638786e-05,
+      "loss": 0.6133,
+      "step": 4071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2542463585102834,
+      "learning_rate": 4.864856526375413e-05,
+      "loss": 1.0254,
+      "step": 4072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1875107708917283,
+      "learning_rate": 4.864721228299045e-05,
+      "loss": 0.8721,
+      "step": 4073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2090769781236412,
+      "learning_rate": 4.864585864413449e-05,
+      "loss": 0.7988,
+      "step": 4074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.245207557004372,
+      "learning_rate": 4.86445043472239e-05,
+      "loss": 0.9785,
+      "step": 4075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.102938292323592,
+      "learning_rate": 4.864314939229638e-05,
+      "loss": 1.3066,
+      "step": 4076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.568835865063382,
+      "learning_rate": 4.864179377938964e-05,
+      "loss": 1.4043,
+      "step": 4077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.614338158346003,
+      "learning_rate": 4.864043750854141e-05,
+      "loss": 1.2285,
+      "step": 4078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4470004267596708,
+      "learning_rate": 4.863908057978942e-05,
+      "loss": 0.9238,
+      "step": 4079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6537493435054333,
+      "learning_rate": 4.8637722993171444e-05,
+      "loss": 1.167,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.573082343759265,
+      "learning_rate": 4.863636474872526e-05,
+      "loss": 1.873,
+      "step": 4081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1558573261400236,
+      "learning_rate": 4.863500584648867e-05,
+      "loss": 1.0273,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2597932717240585,
+      "learning_rate": 4.863364628649948e-05,
+      "loss": 1.2168,
+      "step": 4083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7126754774299133,
+      "learning_rate": 4.863228606879554e-05,
+      "loss": 0.8049,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.887138780070709,
+      "learning_rate": 4.8630925193414706e-05,
+      "loss": 0.8843,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.467701112785452,
+      "learning_rate": 4.862956366039484e-05,
+      "loss": 1.0522,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7110041629389223,
+      "learning_rate": 4.862820146977383e-05,
+      "loss": 0.9307,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.641268499393553,
+      "learning_rate": 4.8626838621589596e-05,
+      "loss": 0.7134,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8235535319667804,
+      "learning_rate": 4.862547511588007e-05,
+      "loss": 0.7982,
+      "step": 4089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.884446918752295,
+      "learning_rate": 4.862411095268318e-05,
+      "loss": 1.1221,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8772884051810577,
+      "learning_rate": 4.86227461320369e-05,
+      "loss": 0.9102,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7177652485681019,
+      "learning_rate": 4.862138065397922e-05,
+      "loss": 0.6694,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8938443226505364,
+      "learning_rate": 4.8620014518548126e-05,
+      "loss": 0.7134,
+      "step": 4093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.028879577654949,
+      "learning_rate": 4.861864772578164e-05,
+      "loss": 0.8975,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.55664084790858,
+      "learning_rate": 4.8617280275717815e-05,
+      "loss": 1.001,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7514668456835247,
+      "learning_rate": 4.8615912168394686e-05,
+      "loss": 1.248,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.184552194683589,
+      "learning_rate": 4.861454340385033e-05,
+      "loss": 1.1055,
+      "step": 4097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.064472526812804,
+      "learning_rate": 4.861317398212285e-05,
+      "loss": 1.0156,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.565051227675218,
+      "learning_rate": 4.861180390325035e-05,
+      "loss": 0.6943,
+      "step": 4099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.596484294195218,
+      "learning_rate": 4.8610433167270964e-05,
+      "loss": 1.0063,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4996093804501758,
+      "learning_rate": 4.860906177422284e-05,
+      "loss": 0.8447,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.44091824637181,
+      "learning_rate": 4.860768972414412e-05,
+      "loss": 1.0674,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.116924529562201,
+      "learning_rate": 4.860631701707302e-05,
+      "loss": 1.1484,
+      "step": 4103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4264599131793037,
+      "learning_rate": 4.860494365304772e-05,
+      "loss": 1.0376,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.792830603468769,
+      "learning_rate": 4.860356963210646e-05,
+      "loss": 1.3223,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4732016101932275,
+      "learning_rate": 4.8602194954287455e-05,
+      "loss": 0.7677,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4685652400118916,
+      "learning_rate": 4.860081961962898e-05,
+      "loss": 1.3809,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.542327977019092,
+      "learning_rate": 4.8599443628169295e-05,
+      "loss": 1.5439,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.8875925047490005,
+      "learning_rate": 4.8598066979946706e-05,
+      "loss": 1.042,
+      "step": 4109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8368670768346262,
+      "learning_rate": 4.859668967499952e-05,
+      "loss": 0.5234,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.628447856871333,
+      "learning_rate": 4.859531171336606e-05,
+      "loss": 1.1841,
+      "step": 4111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.558544796089974,
+      "learning_rate": 4.8593933095084686e-05,
+      "loss": 1.4238,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.691296225016595,
+      "learning_rate": 4.859255382019376e-05,
+      "loss": 1.4238,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1490565292499557,
+      "learning_rate": 4.8591173888731664e-05,
+      "loss": 1.5098,
+      "step": 4114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.99851780703764,
+      "learning_rate": 4.85897933007368e-05,
+      "loss": 0.8228,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.910975805040598,
+      "learning_rate": 4.858841205624759e-05,
+      "loss": 0.5962,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.732728311123784,
+      "learning_rate": 4.858703015530249e-05,
+      "loss": 1.1748,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7748056395228273,
+      "learning_rate": 4.858564759793993e-05,
+      "loss": 0.6953,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3736378628487063,
+      "learning_rate": 4.85842643841984e-05,
+      "loss": 0.7158,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7553000095357323,
+      "learning_rate": 4.8582880514116396e-05,
+      "loss": 0.9199,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8764623026020395,
+      "learning_rate": 4.8581495987732427e-05,
+      "loss": 0.8447,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.541710004354372,
+      "learning_rate": 4.858011080508503e-05,
+      "loss": 0.9912,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2957716695131667,
+      "learning_rate": 4.857872496621274e-05,
+      "loss": 0.6921,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0924918959317917,
+      "learning_rate": 4.857733847115414e-05,
+      "loss": 0.7002,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.229187453967806,
+      "learning_rate": 4.857595131994781e-05,
+      "loss": 0.9121,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0002451984780327,
+      "learning_rate": 4.857456351263234e-05,
+      "loss": 0.9658,
+      "step": 4126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.678607197455199,
+      "learning_rate": 4.8573175049246376e-05,
+      "loss": 1.3374,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2808002319204705,
+      "learning_rate": 4.857178592982854e-05,
+      "loss": 1.0039,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6077980255950353,
+      "learning_rate": 4.857039615441751e-05,
+      "loss": 1.2529,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3899658890209277,
+      "learning_rate": 4.856900572305194e-05,
+      "loss": 1.4121,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.556902750331365,
+      "learning_rate": 4.8567614635770534e-05,
+      "loss": 0.8833,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4985834647238585,
+      "learning_rate": 4.856622289261201e-05,
+      "loss": 1.3438,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.2539019802062255,
+      "learning_rate": 4.85648304936151e-05,
+      "loss": 1.3594,
+      "step": 4133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7674899952060663,
+      "learning_rate": 4.856343743881855e-05,
+      "loss": 1.3164,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.461589705428432,
+      "learning_rate": 4.8562043728261125e-05,
+      "loss": 1.2939,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.902101160518115,
+      "learning_rate": 4.8560649361981615e-05,
+      "loss": 1.4082,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5114126782809767,
+      "learning_rate": 4.855925434001883e-05,
+      "loss": 0.9531,
+      "step": 4137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.004932757841109,
+      "learning_rate": 4.855785866241158e-05,
+      "loss": 1.3223,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4802327682101075,
+      "learning_rate": 4.855646232919872e-05,
+      "loss": 1.0312,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.201896618024175,
+      "learning_rate": 4.85550653404191e-05,
+      "loss": 0.9092,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.853568931045135,
+      "learning_rate": 4.8553667696111604e-05,
+      "loss": 0.75,
+      "step": 4141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.60833953527125,
+      "learning_rate": 4.855226939631512e-05,
+      "loss": 1.9199,
+      "step": 4142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.147699265248903,
+      "learning_rate": 4.8550870441068576e-05,
+      "loss": 0.9248,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5824397560452186,
+      "learning_rate": 4.8549470830410886e-05,
+      "loss": 0.7998,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.883479321709439,
+      "learning_rate": 4.8548070564381016e-05,
+      "loss": 1.3662,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.286047037013384,
+      "learning_rate": 4.854666964301793e-05,
+      "loss": 0.9492,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.628433162298763,
+      "learning_rate": 4.85452680663606e-05,
+      "loss": 1.7871,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9166323705200874,
+      "learning_rate": 4.8543865834448056e-05,
+      "loss": 0.7959,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.218728667253336,
+      "learning_rate": 4.854246294731931e-05,
+      "loss": 0.9795,
+      "step": 4149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.708091715280855,
+      "learning_rate": 4.85410594050134e-05,
+      "loss": 1.3438,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1143544260746783,
+      "learning_rate": 4.8539655207569404e-05,
+      "loss": 1.1631,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.042648435684581,
+      "learning_rate": 4.853825035502637e-05,
+      "loss": 1.0752,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2422792635226725,
+      "learning_rate": 4.8536844847423415e-05,
+      "loss": 1.1738,
+      "step": 4153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.73083062950887,
+      "learning_rate": 4.853543868479965e-05,
+      "loss": 0.7031,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.043545211016585,
+      "learning_rate": 4.853403186719421e-05,
+      "loss": 1.2998,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.240783681635475,
+      "learning_rate": 4.853262439464624e-05,
+      "loss": 1.5449,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.12543133619865,
+      "learning_rate": 4.853121626719492e-05,
+      "loss": 1.6338,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.240367083954752,
+      "learning_rate": 4.852980748487943e-05,
+      "loss": 0.7656,
+      "step": 4158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.406807049403781,
+      "learning_rate": 4.8528398047738976e-05,
+      "loss": 1.0605,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.1099746772153605,
+      "learning_rate": 4.852698795581278e-05,
+      "loss": 1.0107,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.964776280021681,
+      "learning_rate": 4.852557720914009e-05,
+      "loss": 2.1758,
+      "step": 4161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.007653067765755,
+      "learning_rate": 4.852416580776016e-05,
+      "loss": 1.3799,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.07961132105545,
+      "learning_rate": 4.852275375171227e-05,
+      "loss": 0.6709,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8423428250275076,
+      "learning_rate": 4.852134104103573e-05,
+      "loss": 0.6602,
+      "step": 4164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.92421030182043,
+      "learning_rate": 4.8519927675769834e-05,
+      "loss": 1.2139,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.110844250617366,
+      "learning_rate": 4.851851365595393e-05,
+      "loss": 0.8291,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.571272743135641,
+      "learning_rate": 4.851709898162736e-05,
+      "loss": 0.8252,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7556804018388081,
+      "learning_rate": 4.85156836528295e-05,
+      "loss": 0.6505,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.551487122524327,
+      "learning_rate": 4.8514267669599745e-05,
+      "loss": 1.1582,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.689747182705621,
+      "learning_rate": 4.851285103197749e-05,
+      "loss": 1.2656,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.385162495449672,
+      "learning_rate": 4.8511433740002164e-05,
+      "loss": 0.9326,
+      "step": 4171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4229978347847254,
+      "learning_rate": 4.85100157937132e-05,
+      "loss": 0.8975,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.527630479050376,
+      "learning_rate": 4.850859719315008e-05,
+      "loss": 0.3423,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6076745986191,
+      "learning_rate": 4.8507177938352266e-05,
+      "loss": 1.3037,
+      "step": 4174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0977218255393515,
+      "learning_rate": 4.850575802935926e-05,
+      "loss": 1.2461,
+      "step": 4175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.961375203685281,
+      "learning_rate": 4.850433746621057e-05,
+      "loss": 1.6191,
+      "step": 4176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.7126727562448725,
+      "learning_rate": 4.850291624894574e-05,
+      "loss": 1.8652,
+      "step": 4177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.1361875544215305,
+      "learning_rate": 4.850149437760432e-05,
+      "loss": 1.334,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.82369158667085,
+      "learning_rate": 4.850007185222588e-05,
+      "loss": 1.5801,
+      "step": 4179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.53353420849854,
+      "learning_rate": 4.849864867285001e-05,
+      "loss": 1.2734,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.224274894540963,
+      "learning_rate": 4.849722483951631e-05,
+      "loss": 1.2461,
+      "step": 4181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.954770639200012,
+      "learning_rate": 4.849580035226441e-05,
+      "loss": 0.8765,
+      "step": 4182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3146718111017663,
+      "learning_rate": 4.849437521113395e-05,
+      "loss": 0.5166,
+      "step": 4183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0406438899662867,
+      "learning_rate": 4.84929494161646e-05,
+      "loss": 1.2617,
+      "step": 4184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9804994479175964,
+      "learning_rate": 4.849152296739602e-05,
+      "loss": 0.9269,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.690443346286379,
+      "learning_rate": 4.849009586486792e-05,
+      "loss": 1.0078,
+      "step": 4186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2895254246403254,
+      "learning_rate": 4.848866810862003e-05,
+      "loss": 0.9995,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.671002462809066,
+      "learning_rate": 4.848723969869205e-05,
+      "loss": 1.4238,
+      "step": 4188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4680785482942382,
+      "learning_rate": 4.8485810635123766e-05,
+      "loss": 1.166,
+      "step": 4189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.807308220412938,
+      "learning_rate": 4.848438091795492e-05,
+      "loss": 1.0059,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.079069196081474,
+      "learning_rate": 4.848295054722532e-05,
+      "loss": 0.8691,
+      "step": 4191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3668837266450944,
+      "learning_rate": 4.848151952297477e-05,
+      "loss": 0.7002,
+      "step": 4192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3711864572717443,
+      "learning_rate": 4.848008784524308e-05,
+      "loss": 0.9346,
+      "step": 4193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.185359452489332,
+      "learning_rate": 4.8478655514070114e-05,
+      "loss": 0.854,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1532171177086745,
+      "learning_rate": 4.847722252949572e-05,
+      "loss": 1.012,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.568864223047582,
+      "learning_rate": 4.8475788891559783e-05,
+      "loss": 0.9075,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.7729346607395415,
+      "learning_rate": 4.847435460030219e-05,
+      "loss": 1.0762,
+      "step": 4197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.604481491967164,
+      "learning_rate": 4.847291965576287e-05,
+      "loss": 2.1953,
+      "step": 4198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4332857134828254,
+      "learning_rate": 4.847148405798175e-05,
+      "loss": 0.7627,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.828678477382285,
+      "learning_rate": 4.847004780699879e-05,
+      "loss": 1.2852,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0891987625346804,
+      "learning_rate": 4.846861090285395e-05,
+      "loss": 0.7866,
+      "step": 4201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1389936319014993,
+      "learning_rate": 4.846717334558722e-05,
+      "loss": 0.958,
+      "step": 4202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.020857033996782,
+      "learning_rate": 4.846573513523862e-05,
+      "loss": 0.9663,
+      "step": 4203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.687997170246553,
+      "learning_rate": 4.8464296271848155e-05,
+      "loss": 1.1064,
+      "step": 4204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.00387918671936,
+      "learning_rate": 4.846285675545588e-05,
+      "loss": 1.4316,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.24148559419282,
+      "learning_rate": 4.8461416586101846e-05,
+      "loss": 1.7344,
+      "step": 4206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.1421872052998765,
+      "learning_rate": 4.845997576382615e-05,
+      "loss": 1.668,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5449550994913417,
+      "learning_rate": 4.845853428866887e-05,
+      "loss": 0.6445,
+      "step": 4208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8419686911118083,
+      "learning_rate": 4.8457092160670134e-05,
+      "loss": 0.7959,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.763833451686865,
+      "learning_rate": 4.845564937987008e-05,
+      "loss": 1.3857,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5906357970892784,
+      "learning_rate": 4.845420594630884e-05,
+      "loss": 1.0879,
+      "step": 4211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.155978804919648,
+      "learning_rate": 4.845276186002661e-05,
+      "loss": 1.4053,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.052994062873145,
+      "learning_rate": 4.8451317121063554e-05,
+      "loss": 1.1445,
+      "step": 4213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.023083361993488,
+      "learning_rate": 4.8449871729459894e-05,
+      "loss": 1.3359,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7563081454380027,
+      "learning_rate": 4.844842568525585e-05,
+      "loss": 1.2637,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.147737562502652,
+      "learning_rate": 4.844697898849166e-05,
+      "loss": 1.4297,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.626233283583583,
+      "learning_rate": 4.84455316392076e-05,
+      "loss": 1.1797,
+      "step": 4217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.039643537381598,
+      "learning_rate": 4.8444083637443934e-05,
+      "loss": 1.332,
+      "step": 4218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.469938332397227,
+      "learning_rate": 4.8442634983240964e-05,
+      "loss": 1.0811,
+      "step": 4219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.887195224406215,
+      "learning_rate": 4.8441185676639e-05,
+      "loss": 0.9697,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2976109370199773,
+      "learning_rate": 4.843973571767839e-05,
+      "loss": 1.1855,
+      "step": 4221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.02339207814718,
+      "learning_rate": 4.8438285106399473e-05,
+      "loss": 1.0781,
+      "step": 4222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8695941256816284,
+      "learning_rate": 4.8436833842842635e-05,
+      "loss": 1.043,
+      "step": 4223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.167437444639528,
+      "learning_rate": 4.843538192704824e-05,
+      "loss": 0.6914,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4884895465087844,
+      "learning_rate": 4.843392935905671e-05,
+      "loss": 0.5977,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.910514462665573,
+      "learning_rate": 4.843247613890846e-05,
+      "loss": 1.4082,
+      "step": 4226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.082241140811397,
+      "learning_rate": 4.8431022266643943e-05,
+      "loss": 1.0684,
+      "step": 4227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9217102502513197,
+      "learning_rate": 4.8429567742303616e-05,
+      "loss": 1.2891,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.936963296112153,
+      "learning_rate": 4.842811256592796e-05,
+      "loss": 0.9321,
+      "step": 4229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.763429535394131,
+      "learning_rate": 4.8426656737557464e-05,
+      "loss": 1.0801,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.550952596239052,
+      "learning_rate": 4.842520025723265e-05,
+      "loss": 1.5295,
+      "step": 4231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.437230070583192,
+      "learning_rate": 4.842374312499405e-05,
+      "loss": 1.0723,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0505391877231176,
+      "learning_rate": 4.8422285340882214e-05,
+      "loss": 1.3184,
+      "step": 4233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8010615550382902,
+      "learning_rate": 4.842082690493771e-05,
+      "loss": 1.1094,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.889175520025198,
+      "learning_rate": 4.841936781720112e-05,
+      "loss": 1.1436,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.8380743429993585,
+      "learning_rate": 4.841790807771307e-05,
+      "loss": 1.1084,
+      "step": 4236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7613847198708976,
+      "learning_rate": 4.8416447686514174e-05,
+      "loss": 1.5684,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.394099692072489,
+      "learning_rate": 4.841498664364507e-05,
+      "loss": 0.5352,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.808517688469366,
+      "learning_rate": 4.8413524949146414e-05,
+      "loss": 0.7832,
+      "step": 4239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7726869239088527,
+      "learning_rate": 4.841206260305889e-05,
+      "loss": 1.5312,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.770886587916067,
+      "learning_rate": 4.841059960542319e-05,
+      "loss": 2.082,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0852020546204866,
+      "learning_rate": 4.840913595628005e-05,
+      "loss": 1.4512,
+      "step": 4242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.371585801176358,
+      "learning_rate": 4.8407671655670174e-05,
+      "loss": 1.2012,
+      "step": 4243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.334714843851302,
+      "learning_rate": 4.840620670363433e-05,
+      "loss": 0.8857,
+      "step": 4244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.487314848023515,
+      "learning_rate": 4.840474110021328e-05,
+      "loss": 0.7256,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.495107322149994,
+      "learning_rate": 4.840327484544781e-05,
+      "loss": 0.7312,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.266037686350223,
+      "learning_rate": 4.840180793937873e-05,
+      "loss": 0.369,
+      "step": 4247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2513887569995643,
+      "learning_rate": 4.8400340382046866e-05,
+      "loss": 1.4551,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.581861989624073,
+      "learning_rate": 4.8398872173493054e-05,
+      "loss": 1.0664,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.284571698405154,
+      "learning_rate": 4.8397403313758154e-05,
+      "loss": 1.0195,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.757773550285153,
+      "learning_rate": 4.839593380288304e-05,
+      "loss": 2.0195,
+      "step": 4251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.763513230531845,
+      "learning_rate": 4.839446364090863e-05,
+      "loss": 1.0535,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6101866961532916,
+      "learning_rate": 4.839299282787581e-05,
+      "loss": 1.166,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9085287638158934,
+      "learning_rate": 4.8391521363825516e-05,
+      "loss": 0.8159,
+      "step": 4254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0428893014038723,
+      "learning_rate": 4.8390049248798706e-05,
+      "loss": 0.9541,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8448557956935208,
+      "learning_rate": 4.838857648283635e-05,
+      "loss": 0.8086,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3008529574192784,
+      "learning_rate": 4.838710306597944e-05,
+      "loss": 0.957,
+      "step": 4257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3200471600142019,
+      "learning_rate": 4.838562899826897e-05,
+      "loss": 0.7646,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.80824339144369,
+      "learning_rate": 4.838415427974596e-05,
+      "loss": 1.2939,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1546256540630853,
+      "learning_rate": 4.8382678910451456e-05,
+      "loss": 1.4688,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9199726004518776,
+      "learning_rate": 4.838120289042652e-05,
+      "loss": 1.2705,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.693512483954558,
+      "learning_rate": 4.837972621971223e-05,
+      "loss": 0.9023,
+      "step": 4262
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.476234675986296,
+      "learning_rate": 4.837824889834967e-05,
+      "loss": 1.3564,
+      "step": 4263
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.675803765333009,
+      "learning_rate": 4.8376770926379963e-05,
+      "loss": 1.4941,
+      "step": 4264
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4827788250793916,
+      "learning_rate": 4.837529230384423e-05,
+      "loss": 1.2578,
+      "step": 4265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.631509339516237,
+      "learning_rate": 4.837381303078364e-05,
+      "loss": 1.2773,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.976385375745429,
+      "learning_rate": 4.837233310723934e-05,
+      "loss": 2.0742,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.605031309909797,
+      "learning_rate": 4.8370852533252536e-05,
+      "loss": 1.2305,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.337732928508088,
+      "learning_rate": 4.836937130886442e-05,
+      "loss": 1.2031,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.675526850503447,
+      "learning_rate": 4.83678894341162e-05,
+      "loss": 0.9937,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1138835788590407,
+      "learning_rate": 4.836640690904914e-05,
+      "loss": 0.9609,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.673597770022173,
+      "learning_rate": 4.8364923733704484e-05,
+      "loss": 1.063,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.802856357591574,
+      "learning_rate": 4.8363439908123517e-05,
+      "loss": 1.1562,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0042112835883086,
+      "learning_rate": 4.836195543234753e-05,
+      "loss": 1.4297,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.419854307011261,
+      "learning_rate": 4.8360470306417825e-05,
+      "loss": 1.2539,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.04430410842651,
+      "learning_rate": 4.835898453037574e-05,
+      "loss": 1.2676,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8922672547874018,
+      "learning_rate": 4.835749810426264e-05,
+      "loss": 0.6808,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.887913897112835,
+      "learning_rate": 4.835601102811986e-05,
+      "loss": 1.2773,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.544999270219815,
+      "learning_rate": 4.835452330198881e-05,
+      "loss": 1.5098,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.169230594342991,
+      "learning_rate": 4.835303492591089e-05,
+      "loss": 1.1318,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0485978952950266,
+      "learning_rate": 4.835154589992751e-05,
+      "loss": 1.1562,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4317803400185696,
+      "learning_rate": 4.835005622408011e-05,
+      "loss": 1.0986,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.085837210341832,
+      "learning_rate": 4.8348565898410155e-05,
+      "loss": 1.4976,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.653044168523762,
+      "learning_rate": 4.834707492295911e-05,
+      "loss": 0.8857,
+      "step": 4284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.51504525384762,
+      "learning_rate": 4.834558329776848e-05,
+      "loss": 1.0947,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8852127608042664,
+      "learning_rate": 4.8344091022879775e-05,
+      "loss": 1.2754,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.3268027368749,
+      "learning_rate": 4.8342598098334516e-05,
+      "loss": 1.1084,
+      "step": 4287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0160566426227455,
+      "learning_rate": 4.834110452417425e-05,
+      "loss": 1.3008,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.353455884191161,
+      "learning_rate": 4.8339610300440555e-05,
+      "loss": 1.085,
+      "step": 4289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3298039029780115,
+      "learning_rate": 4.8338115427175004e-05,
+      "loss": 1.1357,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.647168060616836,
+      "learning_rate": 4.83366199044192e-05,
+      "loss": 1.332,
+      "step": 4291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2126054167553524,
+      "learning_rate": 4.833512373221476e-05,
+      "loss": 1.1309,
+      "step": 4292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.732168386288432,
+      "learning_rate": 4.833362691060334e-05,
+      "loss": 0.9668,
+      "step": 4293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.801289914792163,
+      "learning_rate": 4.833212943962657e-05,
+      "loss": 0.7061,
+      "step": 4294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2722661146218446,
+      "learning_rate": 4.8330631319326135e-05,
+      "loss": 0.8945,
+      "step": 4295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.533233195465177,
+      "learning_rate": 4.832913254974374e-05,
+      "loss": 1.3652,
+      "step": 4296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.091264975124404,
+      "learning_rate": 4.8327633130921076e-05,
+      "loss": 1.3184,
+      "step": 4297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.763728730698016,
+      "learning_rate": 4.8326133062899875e-05,
+      "loss": 1.1826,
+      "step": 4298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0354370392086008,
+      "learning_rate": 4.8324632345721894e-05,
+      "loss": 1.2676,
+      "step": 4299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.599847818469231,
+      "learning_rate": 4.8323130979428886e-05,
+      "loss": 0.8467,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6398376555088814,
+      "learning_rate": 4.8321628964062636e-05,
+      "loss": 0.6675,
+      "step": 4301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.544943868480111,
+      "learning_rate": 4.832012629966495e-05,
+      "loss": 1.5098,
+      "step": 4302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0920113966856406,
+      "learning_rate": 4.8318622986277644e-05,
+      "loss": 1.0625,
+      "step": 4303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.896952473449484,
+      "learning_rate": 4.8317119023942556e-05,
+      "loss": 0.9619,
+      "step": 4304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4161037469087323,
+      "learning_rate": 4.8315614412701523e-05,
+      "loss": 1.0508,
+      "step": 4305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.159529141391315,
+      "learning_rate": 4.831410915259645e-05,
+      "loss": 0.6436,
+      "step": 4306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6796184298437256,
+      "learning_rate": 4.8312603243669204e-05,
+      "loss": 1.1328,
+      "step": 4307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5384542352842545,
+      "learning_rate": 4.83110966859617e-05,
+      "loss": 0.791,
+      "step": 4308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9852999941147242,
+      "learning_rate": 4.830958947951586e-05,
+      "loss": 1.334,
+      "step": 4309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6435314187307113,
+      "learning_rate": 4.830808162437365e-05,
+      "loss": 1.0234,
+      "step": 4310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.767853437148747,
+      "learning_rate": 4.8306573120577006e-05,
+      "loss": 0.9111,
+      "step": 4311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.641934326343417,
+      "learning_rate": 4.830506396816791e-05,
+      "loss": 1.7422,
+      "step": 4312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.951730227210562,
+      "learning_rate": 4.830355416718838e-05,
+      "loss": 1.3086,
+      "step": 4313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3384834627810998,
+      "learning_rate": 4.8302043717680435e-05,
+      "loss": 0.998,
+      "step": 4314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2799277458119254,
+      "learning_rate": 4.830053261968609e-05,
+      "loss": 0.9805,
+      "step": 4315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4569192687990373,
+      "learning_rate": 4.8299020873247405e-05,
+      "loss": 1.1797,
+      "step": 4316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.990979633391164,
+      "learning_rate": 4.829750847840646e-05,
+      "loss": 0.6943,
+      "step": 4317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.181257701903695,
+      "learning_rate": 4.829599543520533e-05,
+      "loss": 1.3145,
+      "step": 4318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.623173352113494,
+      "learning_rate": 4.829448174368614e-05,
+      "loss": 0.7275,
+      "step": 4319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.809650794819231,
+      "learning_rate": 4.8292967403891e-05,
+      "loss": 1.3018,
+      "step": 4320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.6552467033333995,
+      "learning_rate": 4.829145241586207e-05,
+      "loss": 1.3555,
+      "step": 4321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.223722544477232,
+      "learning_rate": 4.828993677964149e-05,
+      "loss": 1.8008,
+      "step": 4322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7259664758187003,
+      "learning_rate": 4.828842049527146e-05,
+      "loss": 1.2041,
+      "step": 4323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.300403316011521,
+      "learning_rate": 4.8286903562794164e-05,
+      "loss": 1.1992,
+      "step": 4324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0112925028701207,
+      "learning_rate": 4.828538598225182e-05,
+      "loss": 1.0508,
+      "step": 4325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.303433928879315,
+      "learning_rate": 4.828386775368666e-05,
+      "loss": 1.0195,
+      "step": 4326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1465403555598193,
+      "learning_rate": 4.8282348877140945e-05,
+      "loss": 0.7148,
+      "step": 4327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.268684156588648,
+      "learning_rate": 4.828082935265694e-05,
+      "loss": 1.8652,
+      "step": 4328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.914266269746757,
+      "learning_rate": 4.827930918027693e-05,
+      "loss": 1.4609,
+      "step": 4329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7352544156076446,
+      "learning_rate": 4.827778836004322e-05,
+      "loss": 0.8848,
+      "step": 4330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.288911677785137,
+      "learning_rate": 4.827626689199814e-05,
+      "loss": 0.9932,
+      "step": 4331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7653414670957224,
+      "learning_rate": 4.827474477618402e-05,
+      "loss": 0.7861,
+      "step": 4332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8138791460661907,
+      "learning_rate": 4.827322201264324e-05,
+      "loss": 1.1033,
+      "step": 4333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.45759203205533,
+      "learning_rate": 4.8271698601418155e-05,
+      "loss": 1.2793,
+      "step": 4334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3642039617860395,
+      "learning_rate": 4.8270174542551186e-05,
+      "loss": 0.8345,
+      "step": 4335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.701674937902925,
+      "learning_rate": 4.826864983608472e-05,
+      "loss": 1.5273,
+      "step": 4336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.219600039610508,
+      "learning_rate": 4.826712448206121e-05,
+      "loss": 1.3652,
+      "step": 4337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.427222654293589,
+      "learning_rate": 4.8265598480523094e-05,
+      "loss": 1.4233,
+      "step": 4338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8220507054846076,
+      "learning_rate": 4.826407183151284e-05,
+      "loss": 0.916,
+      "step": 4339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.0410086788676525,
+      "learning_rate": 4.826254453507294e-05,
+      "loss": 2.2383,
+      "step": 4340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3177742480788273,
+      "learning_rate": 4.82610165912459e-05,
+      "loss": 0.8066,
+      "step": 4341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.850285999478113,
+      "learning_rate": 4.8259488000074236e-05,
+      "loss": 0.8979,
+      "step": 4342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3569645979386875,
+      "learning_rate": 4.825795876160049e-05,
+      "loss": 1.1792,
+      "step": 4343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2118698688327916,
+      "learning_rate": 4.825642887586722e-05,
+      "loss": 1.083,
+      "step": 4344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6814776521797463,
+      "learning_rate": 4.8254898342917e-05,
+      "loss": 0.9648,
+      "step": 4345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1152000879774584,
+      "learning_rate": 4.825336716279243e-05,
+      "loss": 1.0259,
+      "step": 4346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.816291203576268,
+      "learning_rate": 4.825183533553611e-05,
+      "loss": 0.9072,
+      "step": 4347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.162779956089548,
+      "learning_rate": 4.825030286119069e-05,
+      "loss": 1.3867,
+      "step": 4348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.477026477756907,
+      "learning_rate": 4.824876973979879e-05,
+      "loss": 1.167,
+      "step": 4349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6320571495043255,
+      "learning_rate": 4.8247235971403103e-05,
+      "loss": 1.084,
+      "step": 4350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.495690083217572,
+      "learning_rate": 4.8245701556046305e-05,
+      "loss": 1.5371,
+      "step": 4351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.241851715716173,
+      "learning_rate": 4.8244166493771096e-05,
+      "loss": 1.6465,
+      "step": 4352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0402602162170282,
+      "learning_rate": 4.824263078462019e-05,
+      "loss": 0.6387,
+      "step": 4353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.27444452913692,
+      "learning_rate": 4.824109442863634e-05,
+      "loss": 0.8613,
+      "step": 4354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.592769893463904,
+      "learning_rate": 4.8239557425862295e-05,
+      "loss": 1.5645,
+      "step": 4355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2170462715919212,
+      "learning_rate": 4.823801977634082e-05,
+      "loss": 0.6689,
+      "step": 4356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.922907193078594,
+      "learning_rate": 4.823648148011471e-05,
+      "loss": 1.1367,
+      "step": 4357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8358230230475088,
+      "learning_rate": 4.823494253722679e-05,
+      "loss": 0.627,
+      "step": 4358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7341414973155427,
+      "learning_rate": 4.823340294771987e-05,
+      "loss": 0.3596,
+      "step": 4359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5932276555996947,
+      "learning_rate": 4.823186271163681e-05,
+      "loss": 1.043,
+      "step": 4360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.682567336120087,
+      "learning_rate": 4.823032182902047e-05,
+      "loss": 0.6467,
+      "step": 4361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.810928624539471,
+      "learning_rate": 4.8228780299913726e-05,
+      "loss": 1.2949,
+      "step": 4362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.02956601310972,
+      "learning_rate": 4.822723812435949e-05,
+      "loss": 1.6484,
+      "step": 4363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.836622087223227,
+      "learning_rate": 4.822569530240066e-05,
+      "loss": 1.0435,
+      "step": 4364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.475389461508031,
+      "learning_rate": 4.82241518340802e-05,
+      "loss": 1.0317,
+      "step": 4365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.706601599344036,
+      "learning_rate": 4.8222607719441034e-05,
+      "loss": 1.123,
+      "step": 4366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.710318810829355,
+      "learning_rate": 4.822106295852617e-05,
+      "loss": 1.4326,
+      "step": 4367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4474441938106835,
+      "learning_rate": 4.821951755137856e-05,
+      "loss": 1.1074,
+      "step": 4368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8897579506379942,
+      "learning_rate": 4.821797149804122e-05,
+      "loss": 0.665,
+      "step": 4369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7619455176083574,
+      "learning_rate": 4.8216424798557205e-05,
+      "loss": 0.7676,
+      "step": 4370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8274417189553014,
+      "learning_rate": 4.821487745296953e-05,
+      "loss": 1.2148,
+      "step": 4371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.773344113791694,
+      "learning_rate": 4.821332946132127e-05,
+      "loss": 0.9453,
+      "step": 4372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.220265904903411,
+      "learning_rate": 4.82117808236555e-05,
+      "loss": 1.1895,
+      "step": 4373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.55591657580599,
+      "learning_rate": 4.821023154001532e-05,
+      "loss": 0.5791,
+      "step": 4374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.524040356676671,
+      "learning_rate": 4.820868161044384e-05,
+      "loss": 0.9062,
+      "step": 4375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1637286728313807,
+      "learning_rate": 4.820713103498421e-05,
+      "loss": 1.4854,
+      "step": 4376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.990638500764651,
+      "learning_rate": 4.8205579813679555e-05,
+      "loss": 1.5928,
+      "step": 4377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3045130776820337,
+      "learning_rate": 4.8204027946573076e-05,
+      "loss": 1.2949,
+      "step": 4378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5017102113916003,
+      "learning_rate": 4.8202475433707936e-05,
+      "loss": 1.208,
+      "step": 4379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.617276955968857,
+      "learning_rate": 4.8200922275127355e-05,
+      "loss": 1.0059,
+      "step": 4380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.180347192397241,
+      "learning_rate": 4.819936847087455e-05,
+      "loss": 0.9597,
+      "step": 4381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8444670726783374,
+      "learning_rate": 4.819781402099277e-05,
+      "loss": 0.876,
+      "step": 4382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.749005665354623,
+      "learning_rate": 4.8196258925525264e-05,
+      "loss": 1.0879,
+      "step": 4383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2407233551608594,
+      "learning_rate": 4.819470318451532e-05,
+      "loss": 1.5703,
+      "step": 4384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.834666736877015,
+      "learning_rate": 4.819314679800623e-05,
+      "loss": 2.0273,
+      "step": 4385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.61336042100135,
+      "learning_rate": 4.81915897660413e-05,
+      "loss": 2.2949,
+      "step": 4386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3813807454097202,
+      "learning_rate": 4.8190032088663864e-05,
+      "loss": 0.8306,
+      "step": 4387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.166427463203557,
+      "learning_rate": 4.8188473765917285e-05,
+      "loss": 0.6572,
+      "step": 4388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.191117094209204,
+      "learning_rate": 4.818691479784492e-05,
+      "loss": 1.3252,
+      "step": 4389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1676897298982163,
+      "learning_rate": 4.818535518449016e-05,
+      "loss": 1.248,
+      "step": 4390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3680520773189104,
+      "learning_rate": 4.818379492589639e-05,
+      "loss": 1.0898,
+      "step": 4391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8316412522141143,
+      "learning_rate": 4.818223402210705e-05,
+      "loss": 1.0625,
+      "step": 4392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.842402828193617,
+      "learning_rate": 4.818067247316558e-05,
+      "loss": 1.3496,
+      "step": 4393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.147599857569543,
+      "learning_rate": 4.817911027911544e-05,
+      "loss": 1.5996,
+      "step": 4394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.603170768136113,
+      "learning_rate": 4.817754744000008e-05,
+      "loss": 1.1113,
+      "step": 4395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4551566876911344,
+      "learning_rate": 4.8175983955863014e-05,
+      "loss": 0.9539,
+      "step": 4396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.398647678935289,
+      "learning_rate": 4.817441982674775e-05,
+      "loss": 0.6909,
+      "step": 4397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.880029351816423,
+      "learning_rate": 4.817285505269783e-05,
+      "loss": 2.0059,
+      "step": 4398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6478615428065746,
+      "learning_rate": 4.8171289633756774e-05,
+      "loss": 0.6523,
+      "step": 4399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.70849750370135,
+      "learning_rate": 4.816972356996816e-05,
+      "loss": 1.373,
+      "step": 4400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9301656524446484,
+      "learning_rate": 4.816815686137558e-05,
+      "loss": 1.0767,
+      "step": 4401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1363400083522803,
+      "learning_rate": 4.816658950802262e-05,
+      "loss": 1.4277,
+      "step": 4402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.805466840379769,
+      "learning_rate": 4.81650215099529e-05,
+      "loss": 1.1836,
+      "step": 4403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.300142520137732,
+      "learning_rate": 4.816345286721007e-05,
+      "loss": 0.5894,
+      "step": 4404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4018864110768936,
+      "learning_rate": 4.816188357983778e-05,
+      "loss": 1.2598,
+      "step": 4405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.259263155047934,
+      "learning_rate": 4.816031364787969e-05,
+      "loss": 1.5361,
+      "step": 4406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.736123545839475,
+      "learning_rate": 4.81587430713795e-05,
+      "loss": 1.3906,
+      "step": 4407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3533740005403514,
+      "learning_rate": 4.8157171850380924e-05,
+      "loss": 0.644,
+      "step": 4408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.710092129740723,
+      "learning_rate": 4.815559998492768e-05,
+      "loss": 1.0684,
+      "step": 4409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.570484229879921,
+      "learning_rate": 4.815402747506352e-05,
+      "loss": 1.335,
+      "step": 4410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6656124494822877,
+      "learning_rate": 4.81524543208322e-05,
+      "loss": 1.084,
+      "step": 4411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.632638234791137,
+      "learning_rate": 4.815088052227749e-05,
+      "loss": 0.9629,
+      "step": 4412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.6339499056588895,
+      "learning_rate": 4.814930607944321e-05,
+      "loss": 1.5254,
+      "step": 4413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.478626726695019,
+      "learning_rate": 4.8147730992373164e-05,
+      "loss": 1.2021,
+      "step": 4414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3104641974373696,
+      "learning_rate": 4.8146155261111194e-05,
+      "loss": 1.356,
+      "step": 4415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.3497331337908465,
+      "learning_rate": 4.814457888570114e-05,
+      "loss": 1.4961,
+      "step": 4416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9121766434118321,
+      "learning_rate": 4.8143001866186875e-05,
+      "loss": 1.1299,
+      "step": 4417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5017600402535582,
+      "learning_rate": 4.814142420261229e-05,
+      "loss": 1.2129,
+      "step": 4418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2116378914926234,
+      "learning_rate": 4.813984589502129e-05,
+      "loss": 1.1113,
+      "step": 4419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.684556163824527,
+      "learning_rate": 4.81382669434578e-05,
+      "loss": 0.6909,
+      "step": 4420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6893716549082383,
+      "learning_rate": 4.813668734796577e-05,
+      "loss": 1.457,
+      "step": 4421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.590118298082401,
+      "learning_rate": 4.813510710858914e-05,
+      "loss": 0.9033,
+      "step": 4422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.738027630648697,
+      "learning_rate": 4.81335262253719e-05,
+      "loss": 0.9688,
+      "step": 4423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.111590002422719,
+      "learning_rate": 4.8131944698358044e-05,
+      "loss": 0.6123,
+      "step": 4424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.204317643106308,
+      "learning_rate": 4.813036252759158e-05,
+      "loss": 0.9316,
+      "step": 4425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.341882801149812,
+      "learning_rate": 4.8128779713116545e-05,
+      "loss": 0.7402,
+      "step": 4426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7808845280345516,
+      "learning_rate": 4.812719625497699e-05,
+      "loss": 0.9678,
+      "step": 4427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.962663373989378,
+      "learning_rate": 4.8125612153216976e-05,
+      "loss": 1.3301,
+      "step": 4428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.340841293826618,
+      "learning_rate": 4.812402740788059e-05,
+      "loss": 0.9893,
+      "step": 4429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.409311475110822,
+      "learning_rate": 4.812244201901193e-05,
+      "loss": 1.0654,
+      "step": 4430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.115885002412599,
+      "learning_rate": 4.812085598665513e-05,
+      "loss": 1.502,
+      "step": 4431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7395261956918173,
+      "learning_rate": 4.811926931085432e-05,
+      "loss": 0.8352,
+      "step": 4432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.124858290491201,
+      "learning_rate": 4.811768199165365e-05,
+      "loss": 0.7568,
+      "step": 4433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8446433279263457,
+      "learning_rate": 4.8116094029097305e-05,
+      "loss": 0.6729,
+      "step": 4434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.150363281159926,
+      "learning_rate": 4.8114505423229474e-05,
+      "loss": 0.5259,
+      "step": 4435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.93769333859257,
+      "learning_rate": 4.811291617409437e-05,
+      "loss": 1.625,
+      "step": 4436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.637076000151039,
+      "learning_rate": 4.8111326281736215e-05,
+      "loss": 1.5449,
+      "step": 4437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.210806071243829,
+      "learning_rate": 4.810973574619926e-05,
+      "loss": 1.5918,
+      "step": 4438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6885587691470443,
+      "learning_rate": 4.8108144567527756e-05,
+      "loss": 0.9858,
+      "step": 4439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.526616978102226,
+      "learning_rate": 4.810655274576601e-05,
+      "loss": 0.4937,
+      "step": 4440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.354946198709414,
+      "learning_rate": 4.81049602809583e-05,
+      "loss": 0.6507,
+      "step": 4441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.940555658418585,
+      "learning_rate": 4.810336717314896e-05,
+      "loss": 0.7251,
+      "step": 4442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2815287153762447,
+      "learning_rate": 4.810177342238231e-05,
+      "loss": 0.9072,
+      "step": 4443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1770150766180048,
+      "learning_rate": 4.8100179028702714e-05,
+      "loss": 1.3652,
+      "step": 4444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5334197273728463,
+      "learning_rate": 4.8098583992154534e-05,
+      "loss": 1.1523,
+      "step": 4445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3832945289209375,
+      "learning_rate": 4.8096988312782174e-05,
+      "loss": 0.9473,
+      "step": 4446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.224090812414942,
+      "learning_rate": 4.8095391990630025e-05,
+      "loss": 0.709,
+      "step": 4447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.117374876394707,
+      "learning_rate": 4.8093795025742514e-05,
+      "loss": 1.2871,
+      "step": 4448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7735743527444794,
+      "learning_rate": 4.80921974181641e-05,
+      "loss": 1.6992,
+      "step": 4449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8979792888209293,
+      "learning_rate": 4.809059916793923e-05,
+      "loss": 1.4238,
+      "step": 4450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.951487253581936,
+      "learning_rate": 4.8089000275112384e-05,
+      "loss": 0.792,
+      "step": 4451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7809925897220278,
+      "learning_rate": 4.808740073972806e-05,
+      "loss": 1.2153,
+      "step": 4452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6766340313358694,
+      "learning_rate": 4.8085800561830774e-05,
+      "loss": 1.2812,
+      "step": 4453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3717211124321196,
+      "learning_rate": 4.808419974146505e-05,
+      "loss": 1.0469,
+      "step": 4454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.375661587474769,
+      "learning_rate": 4.808259827867545e-05,
+      "loss": 1.2412,
+      "step": 4455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.15770163945312,
+      "learning_rate": 4.808099617350654e-05,
+      "loss": 1.2246,
+      "step": 4456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3803000605590294,
+      "learning_rate": 4.8079393426002895e-05,
+      "loss": 0.8937,
+      "step": 4457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.372095899241519,
+      "learning_rate": 4.807779003620913e-05,
+      "loss": 0.6992,
+      "step": 4458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.678445511537684,
+      "learning_rate": 4.807618600416986e-05,
+      "loss": 0.7686,
+      "step": 4459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.823943090747327,
+      "learning_rate": 4.807458132992974e-05,
+      "loss": 0.8608,
+      "step": 4460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.818289245634267,
+      "learning_rate": 4.8072976013533404e-05,
+      "loss": 1.031,
+      "step": 4461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3641700775620285,
+      "learning_rate": 4.8071370055025544e-05,
+      "loss": 0.9795,
+      "step": 4462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.278283476740011,
+      "learning_rate": 4.8069763454450846e-05,
+      "loss": 0.8301,
+      "step": 4463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.3124540439188745,
+      "learning_rate": 4.806815621185402e-05,
+      "loss": 0.4673,
+      "step": 4464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.597304975808505,
+      "learning_rate": 4.80665483272798e-05,
+      "loss": 0.8865,
+      "step": 4465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.718119405642901,
+      "learning_rate": 4.806493980077292e-05,
+      "loss": 1.3398,
+      "step": 4466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.743503383451833,
+      "learning_rate": 4.8063330632378176e-05,
+      "loss": 1.585,
+      "step": 4467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5401718767840764,
+      "learning_rate": 4.806172082214032e-05,
+      "loss": 1.3682,
+      "step": 4468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.672293319198456,
+      "learning_rate": 4.806011037010415e-05,
+      "loss": 1.3408,
+      "step": 4469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.335487041609758,
+      "learning_rate": 4.805849927631452e-05,
+      "loss": 0.8052,
+      "step": 4470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7978815698946318,
+      "learning_rate": 4.805688754081622e-05,
+      "loss": 1.0684,
+      "step": 4471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.453795612949073,
+      "learning_rate": 4.805527516365414e-05,
+      "loss": 1.5781,
+      "step": 4472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.992128661380568,
+      "learning_rate": 4.805366214487313e-05,
+      "loss": 1.0664,
+      "step": 4473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.776216338058251,
+      "learning_rate": 4.805204848451809e-05,
+      "loss": 1.2607,
+      "step": 4474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.683210317459966,
+      "learning_rate": 4.805043418263393e-05,
+      "loss": 0.8398,
+      "step": 4475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7312897633466693,
+      "learning_rate": 4.804881923926556e-05,
+      "loss": 1.1562,
+      "step": 4476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3843983860293902,
+      "learning_rate": 4.8047203654457945e-05,
+      "loss": 1.1777,
+      "step": 4477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.924141397556355,
+      "learning_rate": 4.804558742825603e-05,
+      "loss": 1.2285,
+      "step": 4478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8969528632208763,
+      "learning_rate": 4.80439705607048e-05,
+      "loss": 1.209,
+      "step": 4479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7168249709510213,
+      "learning_rate": 4.804235305184925e-05,
+      "loss": 1.1055,
+      "step": 4480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9587358980821454,
+      "learning_rate": 4.804073490173439e-05,
+      "loss": 1.1201,
+      "step": 4481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7608607332138144,
+      "learning_rate": 4.8039116110405265e-05,
+      "loss": 0.5391,
+      "step": 4482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.649755262271707,
+      "learning_rate": 4.803749667790691e-05,
+      "loss": 0.7744,
+      "step": 4483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.69177554888857,
+      "learning_rate": 4.803587660428441e-05,
+      "loss": 0.9609,
+      "step": 4484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5649757520931233,
+      "learning_rate": 4.803425588958283e-05,
+      "loss": 1.0713,
+      "step": 4485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.239140903034594,
+      "learning_rate": 4.80326345338473e-05,
+      "loss": 1.6387,
+      "step": 4486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1562360680171517,
+      "learning_rate": 4.803101253712292e-05,
+      "loss": 0.5896,
+      "step": 4487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.037918387353662,
+      "learning_rate": 4.802938989945484e-05,
+      "loss": 1.0981,
+      "step": 4488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9351329600486693,
+      "learning_rate": 4.8027766620888216e-05,
+      "loss": 0.9216,
+      "step": 4489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6620886718164787,
+      "learning_rate": 4.802614270146821e-05,
+      "loss": 1.104,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2606604406775346,
+      "learning_rate": 4.802451814124004e-05,
+      "loss": 0.7612,
+      "step": 4491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.02899367004791,
+      "learning_rate": 4.80228929402489e-05,
+      "loss": 1.2324,
+      "step": 4492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.6349580661537,
+      "learning_rate": 4.802126709854002e-05,
+      "loss": 1.8301,
+      "step": 4493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.275646961908958,
+      "learning_rate": 4.801964061615865e-05,
+      "loss": 1.3086,
+      "step": 4494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.0505682597202615,
+      "learning_rate": 4.801801349315005e-05,
+      "loss": 1.4277,
+      "step": 4495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.057463104449635,
+      "learning_rate": 4.801638572955951e-05,
+      "loss": 0.9453,
+      "step": 4496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5961552693289742,
+      "learning_rate": 4.801475732543232e-05,
+      "loss": 1.0059,
+      "step": 4497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.321140231575829,
+      "learning_rate": 4.801312828081381e-05,
+      "loss": 1.1726,
+      "step": 4498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.481689286880087,
+      "learning_rate": 4.80114985957493e-05,
+      "loss": 0.7715,
+      "step": 4499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.357457654877331,
+      "learning_rate": 4.8009868270284156e-05,
+      "loss": 0.924,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5518481641913873,
+      "learning_rate": 4.800823730446374e-05,
+      "loss": 1.5391,
+      "step": 4501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.483129029186421,
+      "learning_rate": 4.8006605698333456e-05,
+      "loss": 1.4834,
+      "step": 4502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8460736261111232,
+      "learning_rate": 4.8004973451938696e-05,
+      "loss": 0.3765,
+      "step": 4503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.932310636726268,
+      "learning_rate": 4.80033405653249e-05,
+      "loss": 0.9746,
+      "step": 4504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1445210218619932,
+      "learning_rate": 4.800170703853748e-05,
+      "loss": 1.1709,
+      "step": 4505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.580435212292473,
+      "learning_rate": 4.800007287162193e-05,
+      "loss": 1.4844,
+      "step": 4506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.520824356187384,
+      "learning_rate": 4.7998438064623716e-05,
+      "loss": 0.8584,
+      "step": 4507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.92941742416446,
+      "learning_rate": 4.799680261758833e-05,
+      "loss": 1.4453,
+      "step": 4508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.7736826332694875,
+      "learning_rate": 4.79951665305613e-05,
+      "loss": 1.7227,
+      "step": 4509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.477265363527987,
+      "learning_rate": 4.799352980358813e-05,
+      "loss": 1.002,
+      "step": 4510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7174077721801666,
+      "learning_rate": 4.7991892436714394e-05,
+      "loss": 1.0684,
+      "step": 4511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.948450467385162,
+      "learning_rate": 4.799025442998565e-05,
+      "loss": 0.8311,
+      "step": 4512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6677065069159618,
+      "learning_rate": 4.798861578344749e-05,
+      "loss": 0.6604,
+      "step": 4513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.604479554771107,
+      "learning_rate": 4.798697649714551e-05,
+      "loss": 0.8242,
+      "step": 4514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.7975936894782745,
+      "learning_rate": 4.798533657112533e-05,
+      "loss": 1.0391,
+      "step": 4515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.371892163669225,
+      "learning_rate": 4.7983696005432587e-05,
+      "loss": 1.2139,
+      "step": 4516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.739993965114435,
+      "learning_rate": 4.798205480011294e-05,
+      "loss": 0.9521,
+      "step": 4517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.732293908518011,
+      "learning_rate": 4.798041295521207e-05,
+      "loss": 1.2432,
+      "step": 4518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9006160542954422,
+      "learning_rate": 4.797877047077566e-05,
+      "loss": 1.3984,
+      "step": 4519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7270106955564364,
+      "learning_rate": 4.797712734684942e-05,
+      "loss": 0.8081,
+      "step": 4520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.237309695508947,
+      "learning_rate": 4.797548358347909e-05,
+      "loss": 0.4519,
+      "step": 4521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.775704092085503,
+      "learning_rate": 4.79738391807104e-05,
+      "loss": 0.9395,
+      "step": 4522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.646635198013791,
+      "learning_rate": 4.797219413858911e-05,
+      "loss": 2.043,
+      "step": 4523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5521964430882957,
+      "learning_rate": 4.797054845716101e-05,
+      "loss": 0.7812,
+      "step": 4524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9381794245226884,
+      "learning_rate": 4.79689021364719e-05,
+      "loss": 0.9087,
+      "step": 4525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.493166453221778,
+      "learning_rate": 4.7967255176567603e-05,
+      "loss": 0.8796,
+      "step": 4526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7470542562797298,
+      "learning_rate": 4.7965607577493944e-05,
+      "loss": 0.594,
+      "step": 4527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.926376547050874,
+      "learning_rate": 4.796395933929677e-05,
+      "loss": 1.207,
+      "step": 4528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.926941900295121,
+      "learning_rate": 4.7962310462021955e-05,
+      "loss": 1.2129,
+      "step": 4529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.438955494306978,
+      "learning_rate": 4.796066094571538e-05,
+      "loss": 1.4102,
+      "step": 4530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.694414602903133,
+      "learning_rate": 4.795901079042298e-05,
+      "loss": 1.2422,
+      "step": 4531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7954297238837698,
+      "learning_rate": 4.795735999619064e-05,
+      "loss": 1.4668,
+      "step": 4532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5695065312772654,
+      "learning_rate": 4.795570856306432e-05,
+      "loss": 1.4805,
+      "step": 4533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.670694361492886,
+      "learning_rate": 4.795405649108998e-05,
+      "loss": 1.2314,
+      "step": 4534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.04693407700579,
+      "learning_rate": 4.7952403780313585e-05,
+      "loss": 0.9961,
+      "step": 4535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8572638383864515,
+      "learning_rate": 4.795075043078114e-05,
+      "loss": 1.4629,
+      "step": 4536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.673976509366483,
+      "learning_rate": 4.794909644253865e-05,
+      "loss": 1.4834,
+      "step": 4537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7040164959068775,
+      "learning_rate": 4.794744181563216e-05,
+      "loss": 0.7993,
+      "step": 4538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0966930840237663,
+      "learning_rate": 4.79457865501077e-05,
+      "loss": 0.7842,
+      "step": 4539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.30211617516844,
+      "learning_rate": 4.7944130646011344e-05,
+      "loss": 1.3281,
+      "step": 4540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3375944557077193,
+      "learning_rate": 4.7942474103389166e-05,
+      "loss": 1.1797,
+      "step": 4541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.635221471581595,
+      "learning_rate": 4.794081692228728e-05,
+      "loss": 1.0957,
+      "step": 4542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6730506496689883,
+      "learning_rate": 4.7939159102751796e-05,
+      "loss": 1.1543,
+      "step": 4543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.185659138790522,
+      "learning_rate": 4.793750064482885e-05,
+      "loss": 0.9424,
+      "step": 4544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.445000113167505,
+      "learning_rate": 4.79358415485646e-05,
+      "loss": 0.812,
+      "step": 4545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.845920849125002,
+      "learning_rate": 4.793418181400522e-05,
+      "loss": 1.1147,
+      "step": 4546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.212615006900705,
+      "learning_rate": 4.79325214411969e-05,
+      "loss": 0.874,
+      "step": 4547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.712292629708275,
+      "learning_rate": 4.793086043018583e-05,
+      "loss": 0.8291,
+      "step": 4548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.571374781067266,
+      "learning_rate": 4.7929198781018265e-05,
+      "loss": 1.1289,
+      "step": 4549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.467870531485507,
+      "learning_rate": 4.792753649374042e-05,
+      "loss": 1.2402,
+      "step": 4550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.415876033454337,
+      "learning_rate": 4.792587356839858e-05,
+      "loss": 0.8682,
+      "step": 4551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2341066700665797,
+      "learning_rate": 4.7924210005039004e-05,
+      "loss": 1.3276,
+      "step": 4552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.2842913218088885,
+      "learning_rate": 4.7922545803707995e-05,
+      "loss": 0.8398,
+      "step": 4553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.206441213992401,
+      "learning_rate": 4.7920880964451876e-05,
+      "loss": 1.2559,
+      "step": 4554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.6614164837143175,
+      "learning_rate": 4.791921548731696e-05,
+      "loss": 1.2656,
+      "step": 4555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0790474357386053,
+      "learning_rate": 4.7917549372349616e-05,
+      "loss": 1.127,
+      "step": 4556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3922912176509885,
+      "learning_rate": 4.7915882619596196e-05,
+      "loss": 1.2988,
+      "step": 4557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7142496526455697,
+      "learning_rate": 4.79142152291031e-05,
+      "loss": 1.1016,
+      "step": 4558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8576509875359064,
+      "learning_rate": 4.791254720091671e-05,
+      "loss": 0.9375,
+      "step": 4559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9673326022177906,
+      "learning_rate": 4.791087853508347e-05,
+      "loss": 1.04,
+      "step": 4560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6898156768725467,
+      "learning_rate": 4.790920923164981e-05,
+      "loss": 1.1182,
+      "step": 4561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.480907777161352,
+      "learning_rate": 4.7907539290662176e-05,
+      "loss": 1.9414,
+      "step": 4562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.275793036270385,
+      "learning_rate": 4.790586871216705e-05,
+      "loss": 1.6992,
+      "step": 4563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.55806404046729,
+      "learning_rate": 4.790419749621092e-05,
+      "loss": 1.2275,
+      "step": 4564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9377163746460906,
+      "learning_rate": 4.7902525642840306e-05,
+      "loss": 1.2695,
+      "step": 4565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1533383600079863,
+      "learning_rate": 4.790085315210172e-05,
+      "loss": 1.168,
+      "step": 4566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0924302185622956,
+      "learning_rate": 4.7899180024041725e-05,
+      "loss": 1.4004,
+      "step": 4567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.358156532995214,
+      "learning_rate": 4.789750625870686e-05,
+      "loss": 1.1868,
+      "step": 4568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6203246487703797,
+      "learning_rate": 4.7895831856143716e-05,
+      "loss": 1.4326,
+      "step": 4569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.862381179910949,
+      "learning_rate": 4.78941568163989e-05,
+      "loss": 0.6885,
+      "step": 4570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8680622174767783,
+      "learning_rate": 4.7892481139519016e-05,
+      "loss": 0.7822,
+      "step": 4571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.15197161778873,
+      "learning_rate": 4.7890804825550704e-05,
+      "loss": 1.1436,
+      "step": 4572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2765785024067533,
+      "learning_rate": 4.7889127874540616e-05,
+      "loss": 0.7686,
+      "step": 4573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.214992426778062,
+      "learning_rate": 4.7887450286535405e-05,
+      "loss": 1.1865,
+      "step": 4574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.514088840664692,
+      "learning_rate": 4.788577206158178e-05,
+      "loss": 1.6777,
+      "step": 4575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.823343459306881,
+      "learning_rate": 4.788409319972643e-05,
+      "loss": 1.3438,
+      "step": 4576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4682007775799195,
+      "learning_rate": 4.788241370101609e-05,
+      "loss": 1.3027,
+      "step": 4577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.16403158151338,
+      "learning_rate": 4.7880733565497486e-05,
+      "loss": 0.7412,
+      "step": 4578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8263656758169675,
+      "learning_rate": 4.787905279321738e-05,
+      "loss": 1.127,
+      "step": 4579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2087326461863643,
+      "learning_rate": 4.787737138422256e-05,
+      "loss": 1.1201,
+      "step": 4580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.101000347389459,
+      "learning_rate": 4.78756893385598e-05,
+      "loss": 0.8228,
+      "step": 4581
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.791363886849497,
+      "learning_rate": 4.7874006656275916e-05,
+      "loss": 1.7598,
+      "step": 4582
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.417211991167829,
+      "learning_rate": 4.787232333741774e-05,
+      "loss": 0.915,
+      "step": 4583
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.597476350678478,
+      "learning_rate": 4.7870639382032116e-05,
+      "loss": 1.334,
+      "step": 4584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1422879954637426,
+      "learning_rate": 4.786895479016591e-05,
+      "loss": 0.7505,
+      "step": 4585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.178542163825581,
+      "learning_rate": 4.786726956186601e-05,
+      "loss": 0.8486,
+      "step": 4586
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0786230881441488,
+      "learning_rate": 4.786558369717929e-05,
+      "loss": 0.5356,
+      "step": 4587
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.308550993780172,
+      "learning_rate": 4.7863897196152704e-05,
+      "loss": 1.6289,
+      "step": 4588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.950627319967244,
+      "learning_rate": 4.786221005883316e-05,
+      "loss": 1.3867,
+      "step": 4589
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8756888078045209,
+      "learning_rate": 4.7860522285267605e-05,
+      "loss": 1.2324,
+      "step": 4590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.732564447335856,
+      "learning_rate": 4.785883387550303e-05,
+      "loss": 1.6875,
+      "step": 4591
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9429504270958895,
+      "learning_rate": 4.785714482958642e-05,
+      "loss": 1.3926,
+      "step": 4592
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.4863708520630805,
+      "learning_rate": 4.785545514756477e-05,
+      "loss": 1.0288,
+      "step": 4593
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.878666240154386,
+      "learning_rate": 4.78537648294851e-05,
+      "loss": 0.7998,
+      "step": 4594
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.375734500823105,
+      "learning_rate": 4.785207387539447e-05,
+      "loss": 1.1006,
+      "step": 4595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6387244613907854,
+      "learning_rate": 4.7850382285339924e-05,
+      "loss": 1.1919,
+      "step": 4596
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9891015180070595,
+      "learning_rate": 4.784869005936854e-05,
+      "loss": 1.6426,
+      "step": 4597
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.014553633019367,
+      "learning_rate": 4.784699719752741e-05,
+      "loss": 1.4629,
+      "step": 4598
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.82001493206323,
+      "learning_rate": 4.784530369986365e-05,
+      "loss": 1.0146,
+      "step": 4599
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.848874850447661,
+      "learning_rate": 4.78436095664244e-05,
+      "loss": 1.0215,
+      "step": 4600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.797565992870687,
+      "learning_rate": 4.7841914797256783e-05,
+      "loss": 1.2109,
+      "step": 4601
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.6172298423000315,
+      "learning_rate": 4.7840219392407984e-05,
+      "loss": 0.8911,
+      "step": 4602
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.902076174858142,
+      "learning_rate": 4.783852335192517e-05,
+      "loss": 1.0479,
+      "step": 4603
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.320827883943676,
+      "learning_rate": 4.783682667585555e-05,
+      "loss": 1.0928,
+      "step": 4604
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.766718298175594,
+      "learning_rate": 4.7835129364246344e-05,
+      "loss": 1.3462,
+      "step": 4605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6672529699310836,
+      "learning_rate": 4.783343141714478e-05,
+      "loss": 0.385,
+      "step": 4606
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.20958241179716,
+      "learning_rate": 4.7831732834598116e-05,
+      "loss": 0.8174,
+      "step": 4607
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8095271292172455,
+      "learning_rate": 4.7830033616653624e-05,
+      "loss": 1.0996,
+      "step": 4608
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8728243603164516,
+      "learning_rate": 4.782833376335859e-05,
+      "loss": 0.7842,
+      "step": 4609
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.923408196839368,
+      "learning_rate": 4.782663327476032e-05,
+      "loss": 1.7617,
+      "step": 4610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.567389034410253,
+      "learning_rate": 4.782493215090614e-05,
+      "loss": 0.8662,
+      "step": 4611
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8478872471903127,
+      "learning_rate": 4.782323039184339e-05,
+      "loss": 1.3105,
+      "step": 4612
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.088122585071865,
+      "learning_rate": 4.782152799761943e-05,
+      "loss": 1.0073,
+      "step": 4613
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.464482061695593,
+      "learning_rate": 4.781982496828163e-05,
+      "loss": 0.7461,
+      "step": 4614
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9577374126529954,
+      "learning_rate": 4.7818121303877406e-05,
+      "loss": 0.9072,
+      "step": 4615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.5727625800140945,
+      "learning_rate": 4.7816417004454145e-05,
+      "loss": 2.084,
+      "step": 4616
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3581758192679523,
+      "learning_rate": 4.781471207005929e-05,
+      "loss": 1.0205,
+      "step": 4617
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9842792397769806,
+      "learning_rate": 4.7813006500740285e-05,
+      "loss": 1.2285,
+      "step": 4618
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4417895206968616,
+      "learning_rate": 4.781130029654459e-05,
+      "loss": 1.209,
+      "step": 4619
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5983745508158242,
+      "learning_rate": 4.780959345751971e-05,
+      "loss": 1.2031,
+      "step": 4620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1806791504040284,
+      "learning_rate": 4.780788598371311e-05,
+      "loss": 0.9551,
+      "step": 4621
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6192530437924035,
+      "learning_rate": 4.7806177875172344e-05,
+      "loss": 1.3242,
+      "step": 4622
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3073082310219215,
+      "learning_rate": 4.7804469131944926e-05,
+      "loss": 0.8213,
+      "step": 4623
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.757204952056694,
+      "learning_rate": 4.780275975407841e-05,
+      "loss": 1.5664,
+      "step": 4624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.383459584909194,
+      "learning_rate": 4.780104974162039e-05,
+      "loss": 1.251,
+      "step": 4625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8727799485678625,
+      "learning_rate": 4.779933909461842e-05,
+      "loss": 0.8799,
+      "step": 4626
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.495261134284111,
+      "learning_rate": 4.779762781312013e-05,
+      "loss": 1.2549,
+      "step": 4627
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3583009814232647,
+      "learning_rate": 4.779591589717314e-05,
+      "loss": 0.9004,
+      "step": 4628
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.72981964154787,
+      "learning_rate": 4.779420334682509e-05,
+      "loss": 1.4532,
+      "step": 4629
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.273272216661463,
+      "learning_rate": 4.779249016212365e-05,
+      "loss": 1.2754,
+      "step": 4630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7818315547426575,
+      "learning_rate": 4.779077634311647e-05,
+      "loss": 1.0264,
+      "step": 4631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.459631679377923,
+      "learning_rate": 4.7789061889851275e-05,
+      "loss": 1.2148,
+      "step": 4632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0632731982431167,
+      "learning_rate": 4.7787346802375766e-05,
+      "loss": 1.1074,
+      "step": 4633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.708037775384274,
+      "learning_rate": 4.7785631080737666e-05,
+      "loss": 0.604,
+      "step": 4634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.388578548776049,
+      "learning_rate": 4.778391472498473e-05,
+      "loss": 1.0156,
+      "step": 4635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.341923116123998,
+      "learning_rate": 4.7782197735164726e-05,
+      "loss": 0.8643,
+      "step": 4636
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.282467716055377,
+      "learning_rate": 4.778048011132543e-05,
+      "loss": 1.2588,
+      "step": 4637
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.934843869971582,
+      "learning_rate": 4.777876185351465e-05,
+      "loss": 1.3379,
+      "step": 4638
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.39027142870829,
+      "learning_rate": 4.777704296178019e-05,
+      "loss": 1.208,
+      "step": 4639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8018657973319394,
+      "learning_rate": 4.777532343616991e-05,
+      "loss": 0.9414,
+      "step": 4640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0283916407171363,
+      "learning_rate": 4.777360327673165e-05,
+      "loss": 1.375,
+      "step": 4641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.298255762950546,
+      "learning_rate": 4.7771882483513274e-05,
+      "loss": 1.04,
+      "step": 4642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.90082697641437,
+      "learning_rate": 4.7770161056562684e-05,
+      "loss": 1.3887,
+      "step": 4643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8722774448733994,
+      "learning_rate": 4.776843899592778e-05,
+      "loss": 1.0361,
+      "step": 4644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9170048733183158,
+      "learning_rate": 4.776671630165649e-05,
+      "loss": 0.9336,
+      "step": 4645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.183756753159598,
+      "learning_rate": 4.776499297379675e-05,
+      "loss": 1.252,
+      "step": 4646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.506653993408513,
+      "learning_rate": 4.776326901239653e-05,
+      "loss": 1.8066,
+      "step": 4647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.644121518163122,
+      "learning_rate": 4.776154441750379e-05,
+      "loss": 1.1982,
+      "step": 4648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3965332308519005,
+      "learning_rate": 4.775981918916654e-05,
+      "loss": 0.5596,
+      "step": 4649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.137971734682455,
+      "learning_rate": 4.775809332743279e-05,
+      "loss": 0.958,
+      "step": 4650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0006433432903705,
+      "learning_rate": 4.775636683235056e-05,
+      "loss": 0.7141,
+      "step": 4651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8697617103098323,
+      "learning_rate": 4.7754639703967915e-05,
+      "loss": 1.1992,
+      "step": 4652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.171836367081774,
+      "learning_rate": 4.775291194233291e-05,
+      "loss": 1.418,
+      "step": 4653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.611635849802285,
+      "learning_rate": 4.775118354749362e-05,
+      "loss": 0.8994,
+      "step": 4654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4109446519819517,
+      "learning_rate": 4.774945451949816e-05,
+      "loss": 1.1309,
+      "step": 4655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9269262782526682,
+      "learning_rate": 4.7747724858394636e-05,
+      "loss": 0.8745,
+      "step": 4656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.370697606129883,
+      "learning_rate": 4.77459945642312e-05,
+      "loss": 1.1875,
+      "step": 4657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.270490696340055,
+      "learning_rate": 4.774426363705599e-05,
+      "loss": 1.248,
+      "step": 4658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.965031481277541,
+      "learning_rate": 4.774253207691718e-05,
+      "loss": 1.168,
+      "step": 4659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.689987982684098,
+      "learning_rate": 4.774079988386296e-05,
+      "loss": 1.1631,
+      "step": 4660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4656067774631523,
+      "learning_rate": 4.773906705794155e-05,
+      "loss": 0.9658,
+      "step": 4661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.139857458075181,
+      "learning_rate": 4.773733359920114e-05,
+      "loss": 1.0205,
+      "step": 4662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7240132216302126,
+      "learning_rate": 4.773559950769e-05,
+      "loss": 0.9944,
+      "step": 4663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.000614119116979,
+      "learning_rate": 4.773386478345639e-05,
+      "loss": 0.9521,
+      "step": 4664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4504323383261077,
+      "learning_rate": 4.773212942654857e-05,
+      "loss": 1.0234,
+      "step": 4665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.740306478685461,
+      "learning_rate": 4.7730393437014845e-05,
+      "loss": 0.8383,
+      "step": 4666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.655999264001806,
+      "learning_rate": 4.772865681490352e-05,
+      "loss": 0.8477,
+      "step": 4667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4043048080697718,
+      "learning_rate": 4.7726919560262934e-05,
+      "loss": 1.373,
+      "step": 4668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.297628992587651,
+      "learning_rate": 4.772518167314143e-05,
+      "loss": 1.4297,
+      "step": 4669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.265419547448967,
+      "learning_rate": 4.7723443153587364e-05,
+      "loss": 0.5518,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.145611908844686,
+      "learning_rate": 4.772170400164913e-05,
+      "loss": 0.9053,
+      "step": 4671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1945187908836097,
+      "learning_rate": 4.771996421737512e-05,
+      "loss": 0.8438,
+      "step": 4672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0523950820852037,
+      "learning_rate": 4.7718223800813764e-05,
+      "loss": 1.4766,
+      "step": 4673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.401602097670432,
+      "learning_rate": 4.771648275201348e-05,
+      "loss": 0.8398,
+      "step": 4674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.786055167678018,
+      "learning_rate": 4.771474107102274e-05,
+      "loss": 1.627,
+      "step": 4675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.24172881240253,
+      "learning_rate": 4.771299875788999e-05,
+      "loss": 2.2969,
+      "step": 4676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7654413835387492,
+      "learning_rate": 4.7711255812663734e-05,
+      "loss": 1.4297,
+      "step": 4677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.922886905423199,
+      "learning_rate": 4.7709512235392476e-05,
+      "loss": 1.207,
+      "step": 4678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.129276948310046,
+      "learning_rate": 4.770776802612474e-05,
+      "loss": 1.1692,
+      "step": 4679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2298919947517764,
+      "learning_rate": 4.7706023184909064e-05,
+      "loss": 1.3809,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.934900618301969,
+      "learning_rate": 4.770427771179401e-05,
+      "loss": 1.7227,
+      "step": 4681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6457094078377987,
+      "learning_rate": 4.7702531606828136e-05,
+      "loss": 1.584,
+      "step": 4682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.795205585203516,
+      "learning_rate": 4.7700784870060065e-05,
+      "loss": 0.6501,
+      "step": 4683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.9089653925468575,
+      "learning_rate": 4.769903750153838e-05,
+      "loss": 1.0366,
+      "step": 4684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1070099181388535,
+      "learning_rate": 4.769728950131173e-05,
+      "loss": 1.2148,
+      "step": 4685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.703335023958159,
+      "learning_rate": 4.7695540869428756e-05,
+      "loss": 1.0469,
+      "step": 4686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.653083697584702,
+      "learning_rate": 4.7693791605938107e-05,
+      "loss": 0.9219,
+      "step": 4687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.165704285424437,
+      "learning_rate": 4.769204171088849e-05,
+      "loss": 1.043,
+      "step": 4688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.104712922844751,
+      "learning_rate": 4.769029118432858e-05,
+      "loss": 1.2549,
+      "step": 4689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.513761989223303,
+      "learning_rate": 4.768854002630711e-05,
+      "loss": 1.292,
+      "step": 4690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9505702679839767,
+      "learning_rate": 4.7686788236872804e-05,
+      "loss": 1.0996,
+      "step": 4691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6987583413642295,
+      "learning_rate": 4.768503581607442e-05,
+      "loss": 1.2344,
+      "step": 4692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5954735644825795,
+      "learning_rate": 4.768328276396072e-05,
+      "loss": 0.9062,
+      "step": 4693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4811154844638974,
+      "learning_rate": 4.76815290805805e-05,
+      "loss": 1.5566,
+      "step": 4694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4603300767908105,
+      "learning_rate": 4.767977476598255e-05,
+      "loss": 1.083,
+      "step": 4695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0699915223247505,
+      "learning_rate": 4.767801982021571e-05,
+      "loss": 0.6353,
+      "step": 4696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2097403343442,
+      "learning_rate": 4.767626424332882e-05,
+      "loss": 0.6904,
+      "step": 4697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9733303674189853,
+      "learning_rate": 4.767450803537071e-05,
+      "loss": 1.0986,
+      "step": 4698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.609759669010673,
+      "learning_rate": 4.7672751196390275e-05,
+      "loss": 1.0059,
+      "step": 4699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.895001042551963,
+      "learning_rate": 4.767099372643641e-05,
+      "loss": 1.0605,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.527300761521683,
+      "learning_rate": 4.766923562555802e-05,
+      "loss": 1.4473,
+      "step": 4701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2573869033483587,
+      "learning_rate": 4.7667476893804025e-05,
+      "loss": 0.7461,
+      "step": 4702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.982629548114642,
+      "learning_rate": 4.7665717531223374e-05,
+      "loss": 1.0039,
+      "step": 4703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4133461688773457,
+      "learning_rate": 4.766395753786504e-05,
+      "loss": 0.96,
+      "step": 4704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.514352163684583,
+      "learning_rate": 4.766219691377799e-05,
+      "loss": 1.0938,
+      "step": 4705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1728325797976984,
+      "learning_rate": 4.766043565901123e-05,
+      "loss": 1.3398,
+      "step": 4706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0593700356778175,
+      "learning_rate": 4.765867377361376e-05,
+      "loss": 0.8984,
+      "step": 4707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1738656848739075,
+      "learning_rate": 4.765691125763464e-05,
+      "loss": 1.0833,
+      "step": 4708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2364699139685347,
+      "learning_rate": 4.7655148111122884e-05,
+      "loss": 0.7773,
+      "step": 4709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.495922482723453,
+      "learning_rate": 4.765338433412759e-05,
+      "loss": 0.875,
+      "step": 4710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.165417649301836,
+      "learning_rate": 4.7651619926697834e-05,
+      "loss": 1.0312,
+      "step": 4711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.11651894802415,
+      "learning_rate": 4.764985488888271e-05,
+      "loss": 1.1328,
+      "step": 4712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.48502915615603,
+      "learning_rate": 4.764808922073135e-05,
+      "loss": 0.7449,
+      "step": 4713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2555027373379586,
+      "learning_rate": 4.764632292229289e-05,
+      "loss": 0.8057,
+      "step": 4714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.163093022771086,
+      "learning_rate": 4.7644555993616476e-05,
+      "loss": 1.1719,
+      "step": 4715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.57537696827873,
+      "learning_rate": 4.7642788434751286e-05,
+      "loss": 1.4218,
+      "step": 4716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.117051864815738,
+      "learning_rate": 4.764102024574651e-05,
+      "loss": 1.3828,
+      "step": 4717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2408837971006337,
+      "learning_rate": 4.7639251426651365e-05,
+      "loss": 0.9717,
+      "step": 4718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3202898942764554,
+      "learning_rate": 4.7637481977515067e-05,
+      "loss": 1.2871,
+      "step": 4719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.308500152952938,
+      "learning_rate": 4.7635711898386866e-05,
+      "loss": 0.8491,
+      "step": 4720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.88391832391726,
+      "learning_rate": 4.7633941189316014e-05,
+      "loss": 1.5625,
+      "step": 4721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.3032201588560035,
+      "learning_rate": 4.7632169850351795e-05,
+      "loss": 1.1621,
+      "step": 4722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8732192166053263,
+      "learning_rate": 4.76303978815435e-05,
+      "loss": 1.041,
+      "step": 4723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.406289300565427,
+      "learning_rate": 4.762862528294044e-05,
+      "loss": 1.1504,
+      "step": 4724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.4131934466670595,
+      "learning_rate": 4.762685205459195e-05,
+      "loss": 0.5635,
+      "step": 4725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.821058965345339,
+      "learning_rate": 4.762507819654739e-05,
+      "loss": 1.5391,
+      "step": 4726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.619437191665525,
+      "learning_rate": 4.762330370885611e-05,
+      "loss": 1.3984,
+      "step": 4727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.561763757565643,
+      "learning_rate": 4.762152859156749e-05,
+      "loss": 1.4688,
+      "step": 4728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.044678977493156,
+      "learning_rate": 4.7619752844730945e-05,
+      "loss": 0.7212,
+      "step": 4729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.142950303127046,
+      "learning_rate": 4.761797646839589e-05,
+      "loss": 1.3008,
+      "step": 4730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6314137532435447,
+      "learning_rate": 4.761619946261176e-05,
+      "loss": 1.2617,
+      "step": 4731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.890102979576688,
+      "learning_rate": 4.761442182742799e-05,
+      "loss": 0.8584,
+      "step": 4732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0524485155867276,
+      "learning_rate": 4.7612643562894075e-05,
+      "loss": 1.6992,
+      "step": 4733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.618990252050338,
+      "learning_rate": 4.7610864669059504e-05,
+      "loss": 1.332,
+      "step": 4734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5839751482025366,
+      "learning_rate": 4.7609085145973765e-05,
+      "loss": 1.6562,
+      "step": 4735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.967103813699497,
+      "learning_rate": 4.7607304993686394e-05,
+      "loss": 2.0469,
+      "step": 4736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9433039689750324,
+      "learning_rate": 4.760552421224693e-05,
+      "loss": 1.2363,
+      "step": 4737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.945478619514627,
+      "learning_rate": 4.760374280170492e-05,
+      "loss": 1.0942,
+      "step": 4738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.631549667192214,
+      "learning_rate": 4.7601960762109966e-05,
+      "loss": 0.9502,
+      "step": 4739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.173318919488351,
+      "learning_rate": 4.760017809351163e-05,
+      "loss": 0.8076,
+      "step": 4740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.499083145048961,
+      "learning_rate": 4.759839479595954e-05,
+      "loss": 1.4336,
+      "step": 4741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.944490337277905,
+      "learning_rate": 4.7596610869503334e-05,
+      "loss": 0.9756,
+      "step": 4742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2106939929811698,
+      "learning_rate": 4.759482631419264e-05,
+      "loss": 0.8945,
+      "step": 4743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.696370594302924,
+      "learning_rate": 4.7593041130077124e-05,
+      "loss": 1.001,
+      "step": 4744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0390130752427806,
+      "learning_rate": 4.7591255317206476e-05,
+      "loss": 1.0635,
+      "step": 4745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.694680544990163,
+      "learning_rate": 4.758946887563039e-05,
+      "loss": 0.7159,
+      "step": 4746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.909679567586785,
+      "learning_rate": 4.7587681805398574e-05,
+      "loss": 1.0781,
+      "step": 4747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.200079682546604,
+      "learning_rate": 4.758589410656078e-05,
+      "loss": 1.4302,
+      "step": 4748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.262303886488664,
+      "learning_rate": 4.758410577916674e-05,
+      "loss": 1.1348,
+      "step": 4749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.4459574904392065,
+      "learning_rate": 4.758231682326623e-05,
+      "loss": 0.7383,
+      "step": 4750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6174731543440517,
+      "learning_rate": 4.758052723890904e-05,
+      "loss": 1.3403,
+      "step": 4751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.7504752662905725,
+      "learning_rate": 4.757873702614497e-05,
+      "loss": 1.7812,
+      "step": 4752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8262086915573557,
+      "learning_rate": 4.757694618502384e-05,
+      "loss": 0.947,
+      "step": 4753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5240728308764813,
+      "learning_rate": 4.757515471559548e-05,
+      "loss": 1.1035,
+      "step": 4754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.422653113975636,
+      "learning_rate": 4.7573362617909766e-05,
+      "loss": 1.3818,
+      "step": 4755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0575841257993757,
+      "learning_rate": 4.7571569892016555e-05,
+      "loss": 1.0537,
+      "step": 4756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5925994358092246,
+      "learning_rate": 4.7569776537965743e-05,
+      "loss": 0.8779,
+      "step": 4757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0063600741698635,
+      "learning_rate": 4.756798255580724e-05,
+      "loss": 0.9482,
+      "step": 4758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.19118135928315,
+      "learning_rate": 4.7566187945590967e-05,
+      "loss": 1.1758,
+      "step": 4759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1464269385373025,
+      "learning_rate": 4.756439270736687e-05,
+      "loss": 1.0859,
+      "step": 4760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9229106438715697,
+      "learning_rate": 4.7562596841184917e-05,
+      "loss": 1.2715,
+      "step": 4761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.208946274918721,
+      "learning_rate": 4.756080034709507e-05,
+      "loss": 0.8428,
+      "step": 4762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.875652570884574,
+      "learning_rate": 4.755900322514734e-05,
+      "loss": 0.8496,
+      "step": 4763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.392759038922459,
+      "learning_rate": 4.755720547539173e-05,
+      "loss": 1.0693,
+      "step": 4764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.7353296864450085,
+      "learning_rate": 4.755540709787828e-05,
+      "loss": 1.6953,
+      "step": 4765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2677488584541976,
+      "learning_rate": 4.755360809265702e-05,
+      "loss": 1.0518,
+      "step": 4766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.03761573416001,
+      "learning_rate": 4.755180845977804e-05,
+      "loss": 1.1294,
+      "step": 4767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.365188692868407,
+      "learning_rate": 4.755000819929141e-05,
+      "loss": 1.502,
+      "step": 4768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.888694984905868,
+      "learning_rate": 4.754820731124723e-05,
+      "loss": 1.168,
+      "step": 4769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7541760900877748,
+      "learning_rate": 4.754640579569562e-05,
+      "loss": 1.1113,
+      "step": 4770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.427449323751166,
+      "learning_rate": 4.754460365268671e-05,
+      "loss": 1.3301,
+      "step": 4771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0688900503030405,
+      "learning_rate": 4.7542800882270665e-05,
+      "loss": 1.4492,
+      "step": 4772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2532889523631865,
+      "learning_rate": 4.754099748449765e-05,
+      "loss": 1.0029,
+      "step": 4773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.006384046576428,
+      "learning_rate": 4.753919345941785e-05,
+      "loss": 1.2041,
+      "step": 4774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.213997884162332,
+      "learning_rate": 4.7537388807081477e-05,
+      "loss": 0.54,
+      "step": 4775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.384516544499411,
+      "learning_rate": 4.753558352753874e-05,
+      "loss": 1.0723,
+      "step": 4776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.933805699317468,
+      "learning_rate": 4.7533777620839886e-05,
+      "loss": 1.0908,
+      "step": 4777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.370522645550476,
+      "learning_rate": 4.7531971087035175e-05,
+      "loss": 1.1943,
+      "step": 4778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.999555872785163,
+      "learning_rate": 4.753016392617489e-05,
+      "loss": 1.1035,
+      "step": 4779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7373215291461395,
+      "learning_rate": 4.75283561383093e-05,
+      "loss": 0.8838,
+      "step": 4780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.12191726023491,
+      "learning_rate": 4.752654772348875e-05,
+      "loss": 1.1738,
+      "step": 4781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.492667419090804,
+      "learning_rate": 4.752473868176354e-05,
+      "loss": 1.082,
+      "step": 4782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.577151582924852,
+      "learning_rate": 4.752292901318401e-05,
+      "loss": 1.4043,
+      "step": 4783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6519792181052533,
+      "learning_rate": 4.752111871780055e-05,
+      "loss": 1.2285,
+      "step": 4784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.921844079734581,
+      "learning_rate": 4.751930779566351e-05,
+      "loss": 0.9258,
+      "step": 4785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.789296279851683,
+      "learning_rate": 4.7517496246823317e-05,
+      "loss": 1.3203,
+      "step": 4786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.6685488697780935,
+      "learning_rate": 4.7515684071330365e-05,
+      "loss": 0.9844,
+      "step": 4787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3279698467399275,
+      "learning_rate": 4.751387126923509e-05,
+      "loss": 1.0527,
+      "step": 4788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3989491746182647,
+      "learning_rate": 4.7512057840587946e-05,
+      "loss": 1.1133,
+      "step": 4789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8197325595887244,
+      "learning_rate": 4.751024378543939e-05,
+      "loss": 1.3945,
+      "step": 4790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.954225243472639,
+      "learning_rate": 4.750842910383991e-05,
+      "loss": 1.0625,
+      "step": 4791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9416551010467558,
+      "learning_rate": 4.750661379584003e-05,
+      "loss": 0.7041,
+      "step": 4792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.1603559559615055,
+      "learning_rate": 4.750479786149023e-05,
+      "loss": 1.4355,
+      "step": 4793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0916936957179457,
+      "learning_rate": 4.7502981300841075e-05,
+      "loss": 0.7139,
+      "step": 4794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.430782549718529,
+      "learning_rate": 4.750116411394312e-05,
+      "loss": 1.1426,
+      "step": 4795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.1179617745476635,
+      "learning_rate": 4.749934630084691e-05,
+      "loss": 1.416,
+      "step": 4796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.560895993709327,
+      "learning_rate": 4.749752786160307e-05,
+      "loss": 1.0059,
+      "step": 4797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.478073477480044,
+      "learning_rate": 4.749570879626218e-05,
+      "loss": 0.96,
+      "step": 4798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.318449026855935,
+      "learning_rate": 4.7493889104874875e-05,
+      "loss": 2.7031,
+      "step": 4799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.073874119866312,
+      "learning_rate": 4.749206878749179e-05,
+      "loss": 0.8494,
+      "step": 4800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5414135262772013,
+      "learning_rate": 4.749024784416359e-05,
+      "loss": 1.1445,
+      "step": 4801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4958714249316034,
+      "learning_rate": 4.748842627494095e-05,
+      "loss": 0.875,
+      "step": 4802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.435802150794582,
+      "learning_rate": 4.7486604079874554e-05,
+      "loss": 1.5859,
+      "step": 4803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4285691806237852,
+      "learning_rate": 4.748478125901512e-05,
+      "loss": 1.3848,
+      "step": 4804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.868630126673631,
+      "learning_rate": 4.7482957812413384e-05,
+      "loss": 1.4102,
+      "step": 4805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.977121603750055,
+      "learning_rate": 4.748113374012009e-05,
+      "loss": 1.4512,
+      "step": 4806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.034714921383296,
+      "learning_rate": 4.7479309042185984e-05,
+      "loss": 1.3105,
+      "step": 4807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.539412629855414,
+      "learning_rate": 4.7477483718661866e-05,
+      "loss": 1.043,
+      "step": 4808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4271604883157725,
+      "learning_rate": 4.747565776959852e-05,
+      "loss": 1.2598,
+      "step": 4809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.11189175253274,
+      "learning_rate": 4.747383119504678e-05,
+      "loss": 1.1748,
+      "step": 4810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2796092991949997,
+      "learning_rate": 4.7472003995057454e-05,
+      "loss": 1.2695,
+      "step": 4811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4183021032138328,
+      "learning_rate": 4.7470176169681414e-05,
+      "loss": 1.7285,
+      "step": 4812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6492881568749467,
+      "learning_rate": 4.746834771896952e-05,
+      "loss": 0.6055,
+      "step": 4813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.9905222104387335,
+      "learning_rate": 4.746651864297265e-05,
+      "loss": 1.5435,
+      "step": 4814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.102393888952789,
+      "learning_rate": 4.7464688941741714e-05,
+      "loss": 1.2695,
+      "step": 4815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.821663657012374,
+      "learning_rate": 4.7462858615327634e-05,
+      "loss": 1.3262,
+      "step": 4816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.042986205427294,
+      "learning_rate": 4.746102766378133e-05,
+      "loss": 1.5186,
+      "step": 4817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0326939072494037,
+      "learning_rate": 4.7459196087153785e-05,
+      "loss": 0.5842,
+      "step": 4818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4463116734234482,
+      "learning_rate": 4.745736388549595e-05,
+      "loss": 1.3242,
+      "step": 4819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2890369840162275,
+      "learning_rate": 4.745553105885882e-05,
+      "loss": 1.0566,
+      "step": 4820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.937977982405395,
+      "learning_rate": 4.74536976072934e-05,
+      "loss": 1.1875,
+      "step": 4821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.020242427165021,
+      "learning_rate": 4.7451863530850724e-05,
+      "loss": 1.0361,
+      "step": 4822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.546540028401641,
+      "learning_rate": 4.7450028829581824e-05,
+      "loss": 1.1768,
+      "step": 4823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.264312772528406,
+      "learning_rate": 4.7448193503537755e-05,
+      "loss": 1.5371,
+      "step": 4824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2649755138901013,
+      "learning_rate": 4.7446357552769604e-05,
+      "loss": 1.0264,
+      "step": 4825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9831018042928545,
+      "learning_rate": 4.7444520977328466e-05,
+      "loss": 1.5684,
+      "step": 4826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.280604022811129,
+      "learning_rate": 4.744268377726544e-05,
+      "loss": 0.96,
+      "step": 4827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.131065016071112,
+      "learning_rate": 4.7440845952631655e-05,
+      "loss": 0.6553,
+      "step": 4828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.572861170202398,
+      "learning_rate": 4.743900750347827e-05,
+      "loss": 1.4316,
+      "step": 4829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7554775652137398,
+      "learning_rate": 4.743716842985644e-05,
+      "loss": 0.9346,
+      "step": 4830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.464779425263822,
+      "learning_rate": 4.7435328731817343e-05,
+      "loss": 1.5957,
+      "step": 4831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0583300977585655,
+      "learning_rate": 4.743348840941219e-05,
+      "loss": 0.7451,
+      "step": 4832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.32066131949674,
+      "learning_rate": 4.743164746269218e-05,
+      "loss": 0.3022,
+      "step": 4833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9627573148333348,
+      "learning_rate": 4.742980589170856e-05,
+      "loss": 1.1904,
+      "step": 4834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.045235476787516,
+      "learning_rate": 4.742796369651257e-05,
+      "loss": 0.7178,
+      "step": 4835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4111302415564464,
+      "learning_rate": 4.742612087715547e-05,
+      "loss": 0.5476,
+      "step": 4836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.783282477521647,
+      "learning_rate": 4.742427743368857e-05,
+      "loss": 1.2207,
+      "step": 4837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.319331186459463,
+      "learning_rate": 4.7422433366163146e-05,
+      "loss": 0.4707,
+      "step": 4838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8384460619770335,
+      "learning_rate": 4.7420588674630536e-05,
+      "loss": 0.9375,
+      "step": 4839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1646733162185203,
+      "learning_rate": 4.741874335914207e-05,
+      "loss": 1.5195,
+      "step": 4840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7463137084873748,
+      "learning_rate": 4.741689741974911e-05,
+      "loss": 1.4414,
+      "step": 4841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.23662639603783,
+      "learning_rate": 4.741505085650301e-05,
+      "loss": 1.7344,
+      "step": 4842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0855833745957355,
+      "learning_rate": 4.741320366945518e-05,
+      "loss": 1.5547,
+      "step": 4843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0103771026721833,
+      "learning_rate": 4.7411355858657006e-05,
+      "loss": 0.8848,
+      "step": 4844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.360118319406445,
+      "learning_rate": 4.740950742415993e-05,
+      "loss": 0.6357,
+      "step": 4845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.243424886156799,
+      "learning_rate": 4.740765836601538e-05,
+      "loss": 1.5039,
+      "step": 4846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5748569004332533,
+      "learning_rate": 4.740580868427482e-05,
+      "loss": 1.2715,
+      "step": 4847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.986318225232937,
+      "learning_rate": 4.740395837898973e-05,
+      "loss": 1.1602,
+      "step": 4848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1227445473974944,
+      "learning_rate": 4.7402107450211595e-05,
+      "loss": 0.8594,
+      "step": 4849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6217713254816246,
+      "learning_rate": 4.740025589799193e-05,
+      "loss": 0.917,
+      "step": 4850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.762490564372009,
+      "learning_rate": 4.739840372238226e-05,
+      "loss": 1.3516,
+      "step": 4851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7295619961062845,
+      "learning_rate": 4.7396550923434136e-05,
+      "loss": 1.0127,
+      "step": 4852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.263959601402425,
+      "learning_rate": 4.739469750119912e-05,
+      "loss": 0.9675,
+      "step": 4853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8657586284684817,
+      "learning_rate": 4.7392843455728787e-05,
+      "loss": 0.8877,
+      "step": 4854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.746226764819464,
+      "learning_rate": 4.739098878707474e-05,
+      "loss": 0.916,
+      "step": 4855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5313907160322673,
+      "learning_rate": 4.7389133495288584e-05,
+      "loss": 1.3086,
+      "step": 4856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2553865836613065,
+      "learning_rate": 4.7387277580421965e-05,
+      "loss": 1.3301,
+      "step": 4857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.599703471073616,
+      "learning_rate": 4.7385421042526515e-05,
+      "loss": 0.8955,
+      "step": 4858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6357799978183714,
+      "learning_rate": 4.738356388165392e-05,
+      "loss": 0.8936,
+      "step": 4859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4967630949685433,
+      "learning_rate": 4.738170609785585e-05,
+      "loss": 1.0488,
+      "step": 4860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3507382795590783,
+      "learning_rate": 4.7379847691184006e-05,
+      "loss": 0.6249,
+      "step": 4861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3216356813873844,
+      "learning_rate": 4.737798866169012e-05,
+      "loss": 0.8364,
+      "step": 4862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.000853258225864,
+      "learning_rate": 4.737612900942592e-05,
+      "loss": 0.9355,
+      "step": 4863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5573224373078176,
+      "learning_rate": 4.737426873444315e-05,
+      "loss": 0.8584,
+      "step": 4864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.061832092570745,
+      "learning_rate": 4.73724078367936e-05,
+      "loss": 1.7441,
+      "step": 4865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.722619471034764,
+      "learning_rate": 4.737054631652904e-05,
+      "loss": 0.9512,
+      "step": 4866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.944492764856518,
+      "learning_rate": 4.7368684173701286e-05,
+      "loss": 0.9531,
+      "step": 4867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.8439598776355055,
+      "learning_rate": 4.736682140836216e-05,
+      "loss": 1.291,
+      "step": 4868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.706835730832966,
+      "learning_rate": 4.7364958020563497e-05,
+      "loss": 0.8311,
+      "step": 4869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8503458395739003,
+      "learning_rate": 4.736309401035716e-05,
+      "loss": 0.8091,
+      "step": 4870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.753421263186753,
+      "learning_rate": 4.736122937779501e-05,
+      "loss": 1.4668,
+      "step": 4871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2364622384048953,
+      "learning_rate": 4.7359364122928964e-05,
+      "loss": 1.502,
+      "step": 4872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.53837057542224,
+      "learning_rate": 4.735749824581091e-05,
+      "loss": 1.1191,
+      "step": 4873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0907802871942494,
+      "learning_rate": 4.735563174649278e-05,
+      "loss": 0.5801,
+      "step": 4874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.218418902555506,
+      "learning_rate": 4.735376462502653e-05,
+      "loss": 0.7935,
+      "step": 4875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0938682148243943,
+      "learning_rate": 4.73518968814641e-05,
+      "loss": 1.0483,
+      "step": 4876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.72781994770812,
+      "learning_rate": 4.735002851585748e-05,
+      "loss": 1.209,
+      "step": 4877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.097630648901198,
+      "learning_rate": 4.734815952825867e-05,
+      "loss": 0.8848,
+      "step": 4878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.768992275335995,
+      "learning_rate": 4.734628991871968e-05,
+      "loss": 2.3125,
+      "step": 4879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.34625059880781,
+      "learning_rate": 4.734441968729253e-05,
+      "loss": 1.1992,
+      "step": 4880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6791319869087427,
+      "learning_rate": 4.734254883402928e-05,
+      "loss": 0.876,
+      "step": 4881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.932988537406698,
+      "learning_rate": 4.734067735898199e-05,
+      "loss": 0.7969,
+      "step": 4882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0997988974037822,
+      "learning_rate": 4.7338805262202744e-05,
+      "loss": 1.6562,
+      "step": 4883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.997045856608851,
+      "learning_rate": 4.733693254374365e-05,
+      "loss": 1.5781,
+      "step": 4884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.59134616367864,
+      "learning_rate": 4.733505920365681e-05,
+      "loss": 1.5332,
+      "step": 4885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4120000528743604,
+      "learning_rate": 4.733318524199436e-05,
+      "loss": 0.6382,
+      "step": 4886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.530701586243317,
+      "learning_rate": 4.733131065880846e-05,
+      "loss": 0.7871,
+      "step": 4887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.17154894240612,
+      "learning_rate": 4.732943545415128e-05,
+      "loss": 0.6216,
+      "step": 4888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.999895666941247,
+      "learning_rate": 4.7327559628075003e-05,
+      "loss": 1.377,
+      "step": 4889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.105163665352973,
+      "learning_rate": 4.732568318063183e-05,
+      "loss": 0.7913,
+      "step": 4890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8466284300702607,
+      "learning_rate": 4.732380611187398e-05,
+      "loss": 0.5674,
+      "step": 4891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.665676022335427,
+      "learning_rate": 4.732192842185369e-05,
+      "loss": 1.3594,
+      "step": 4892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2927732579315894,
+      "learning_rate": 4.7320050110623226e-05,
+      "loss": 1.1143,
+      "step": 4893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.98104256921962,
+      "learning_rate": 4.731817117823485e-05,
+      "loss": 1.6758,
+      "step": 4894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.995211360158249,
+      "learning_rate": 4.7316291624740855e-05,
+      "loss": 1.7656,
+      "step": 4895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.456632888109686,
+      "learning_rate": 4.731441145019355e-05,
+      "loss": 1.2246,
+      "step": 4896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.912093491897916,
+      "learning_rate": 4.7312530654645256e-05,
+      "loss": 1.2061,
+      "step": 4897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.03274289706726,
+      "learning_rate": 4.731064923814833e-05,
+      "loss": 1.0884,
+      "step": 4898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0179935305192997,
+      "learning_rate": 4.73087672007551e-05,
+      "loss": 0.8071,
+      "step": 4899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.977932782950945,
+      "learning_rate": 4.730688454251798e-05,
+      "loss": 1.1348,
+      "step": 4900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.295343498348613,
+      "learning_rate": 4.7305001263489335e-05,
+      "loss": 0.8333,
+      "step": 4901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8784474450145945,
+      "learning_rate": 4.730311736372158e-05,
+      "loss": 1.1992,
+      "step": 4902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.3343710132919435,
+      "learning_rate": 4.730123284326715e-05,
+      "loss": 1.3105,
+      "step": 4903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6975672040174326,
+      "learning_rate": 4.72993477021785e-05,
+      "loss": 1.1602,
+      "step": 4904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.012049841703757,
+      "learning_rate": 4.729746194050807e-05,
+      "loss": 1.2812,
+      "step": 4905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5229579599803817,
+      "learning_rate": 4.729557555830836e-05,
+      "loss": 0.9658,
+      "step": 4906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4711718653291057,
+      "learning_rate": 4.7293688555631846e-05,
+      "loss": 1.0078,
+      "step": 4907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.935989990966255,
+      "learning_rate": 4.7291800932531064e-05,
+      "loss": 1.2021,
+      "step": 4908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.160040578107797,
+      "learning_rate": 4.728991268905853e-05,
+      "loss": 0.8721,
+      "step": 4909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.430137221232565,
+      "learning_rate": 4.728802382526681e-05,
+      "loss": 1.1484,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.756100909640482,
+      "learning_rate": 4.728613434120845e-05,
+      "loss": 0.9873,
+      "step": 4911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.583265416472837,
+      "learning_rate": 4.728424423693606e-05,
+      "loss": 0.9131,
+      "step": 4912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.268009178055145,
+      "learning_rate": 4.7282353512502216e-05,
+      "loss": 1.2607,
+      "step": 4913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.428924441001626,
+      "learning_rate": 4.728046216795954e-05,
+      "loss": 0.8594,
+      "step": 4914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.546063973217127,
+      "learning_rate": 4.727857020336068e-05,
+      "loss": 0.856,
+      "step": 4915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5508296664025827,
+      "learning_rate": 4.727667761875828e-05,
+      "loss": 0.8491,
+      "step": 4916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.634481144758702,
+      "learning_rate": 4.7274784414205006e-05,
+      "loss": 2.0332,
+      "step": 4917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.840991987143147,
+      "learning_rate": 4.7272890589753546e-05,
+      "loss": 1.585,
+      "step": 4918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1570193258948644,
+      "learning_rate": 4.7270996145456615e-05,
+      "loss": 0.9756,
+      "step": 4919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.243215872661708,
+      "learning_rate": 4.726910108136693e-05,
+      "loss": 1.2285,
+      "step": 4920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.046022892666187,
+      "learning_rate": 4.726720539753722e-05,
+      "loss": 0.9141,
+      "step": 4921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4208191569750905,
+      "learning_rate": 4.726530909402026e-05,
+      "loss": 0.9629,
+      "step": 4922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9420961176395455,
+      "learning_rate": 4.7263412170868804e-05,
+      "loss": 0.8564,
+      "step": 4923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2521342092506003,
+      "learning_rate": 4.726151462813565e-05,
+      "loss": 1.1914,
+      "step": 4924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.708389981869409,
+      "learning_rate": 4.7259616465873604e-05,
+      "loss": 0.712,
+      "step": 4925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.177664416091825,
+      "learning_rate": 4.725771768413551e-05,
+      "loss": 0.9521,
+      "step": 4926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.108228642364288,
+      "learning_rate": 4.725581828297418e-05,
+      "loss": 0.6598,
+      "step": 4927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.536085807708418,
+      "learning_rate": 4.725391826244249e-05,
+      "loss": 0.9072,
+      "step": 4928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.358016622169346,
+      "learning_rate": 4.7252017622593307e-05,
+      "loss": 1.3564,
+      "step": 4929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.334453200494774,
+      "learning_rate": 4.725011636347954e-05,
+      "loss": 0.6729,
+      "step": 4930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.7421124875013465,
+      "learning_rate": 4.7248214485154097e-05,
+      "loss": 0.6875,
+      "step": 4931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.639750494318075,
+      "learning_rate": 4.724631198766989e-05,
+      "loss": 0.96,
+      "step": 4932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.507048433321335,
+      "learning_rate": 4.724440887107989e-05,
+      "loss": 1.6738,
+      "step": 4933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2931521600678035,
+      "learning_rate": 4.724250513543704e-05,
+      "loss": 1.2227,
+      "step": 4934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.432583921427675,
+      "learning_rate": 4.724060078079432e-05,
+      "loss": 1.541,
+      "step": 4935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.869686791020134,
+      "learning_rate": 4.7238695807204745e-05,
+      "loss": 1.2734,
+      "step": 4936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.635721201634684,
+      "learning_rate": 4.7236790214721315e-05,
+      "loss": 1.1406,
+      "step": 4937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9062545940403877,
+      "learning_rate": 4.723488400339707e-05,
+      "loss": 1.0615,
+      "step": 4938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.98956559113358,
+      "learning_rate": 4.7232977173285055e-05,
+      "loss": 1.0176,
+      "step": 4939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4434930534599704,
+      "learning_rate": 4.723106972443833e-05,
+      "loss": 0.793,
+      "step": 4940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.445698667996508,
+      "learning_rate": 4.722916165690999e-05,
+      "loss": 1.418,
+      "step": 4941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1076325177831854,
+      "learning_rate": 4.722725297075313e-05,
+      "loss": 1.6201,
+      "step": 4942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.359064504629701,
+      "learning_rate": 4.722534366602087e-05,
+      "loss": 1.5771,
+      "step": 4943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.692720888087858,
+      "learning_rate": 4.722343374276634e-05,
+      "loss": 1.3867,
+      "step": 4944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.770860541431298,
+      "learning_rate": 4.722152320104269e-05,
+      "loss": 1.3271,
+      "step": 4945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9969681678585967,
+      "learning_rate": 4.72196120409031e-05,
+      "loss": 1.2188,
+      "step": 4946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4688893532354355,
+      "learning_rate": 4.7217700262400765e-05,
+      "loss": 1.2012,
+      "step": 4947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2953487678221425,
+      "learning_rate": 4.7215787865588865e-05,
+      "loss": 1.3086,
+      "step": 4948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.056793608222699,
+      "learning_rate": 4.721387485052064e-05,
+      "loss": 1.2324,
+      "step": 4949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1127347191241057,
+      "learning_rate": 4.721196121724931e-05,
+      "loss": 1.085,
+      "step": 4950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8110199854058227,
+      "learning_rate": 4.721004696582815e-05,
+      "loss": 1.1602,
+      "step": 4951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8703064130906473,
+      "learning_rate": 4.720813209631042e-05,
+      "loss": 0.7744,
+      "step": 4952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.052730859265948,
+      "learning_rate": 4.720621660874942e-05,
+      "loss": 1.3926,
+      "step": 4953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.271765024915688,
+      "learning_rate": 4.720430050319845e-05,
+      "loss": 1.0063,
+      "step": 4954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0714169417369583,
+      "learning_rate": 4.720238377971084e-05,
+      "loss": 0.9309,
+      "step": 4955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.191813971670112,
+      "learning_rate": 4.720046643833992e-05,
+      "loss": 0.7988,
+      "step": 4956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.315747351434745,
+      "learning_rate": 4.719854847913906e-05,
+      "loss": 1.2754,
+      "step": 4957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7171326134322284,
+      "learning_rate": 4.7196629902161627e-05,
+      "loss": 1.4531,
+      "step": 4958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2512935118383375,
+      "learning_rate": 4.719471070746103e-05,
+      "loss": 1.2871,
+      "step": 4959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1303600701819136,
+      "learning_rate": 4.719279089509067e-05,
+      "loss": 0.855,
+      "step": 4960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.708172190561489,
+      "learning_rate": 4.719087046510396e-05,
+      "loss": 0.5962,
+      "step": 4961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.32451007692149,
+      "learning_rate": 4.718894941755437e-05,
+      "loss": 0.7246,
+      "step": 4962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.745986437159063,
+      "learning_rate": 4.718702775249535e-05,
+      "loss": 1.5273,
+      "step": 4963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2349686377300735,
+      "learning_rate": 4.718510546998038e-05,
+      "loss": 1.0332,
+      "step": 4964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.402416832816406,
+      "learning_rate": 4.7183182570062955e-05,
+      "loss": 1.1631,
+      "step": 4965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.330775550650681,
+      "learning_rate": 4.7181259052796597e-05,
+      "loss": 0.8418,
+      "step": 4966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8505471289967033,
+      "learning_rate": 4.717933491823482e-05,
+      "loss": 1.1211,
+      "step": 4967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.702952320563512,
+      "learning_rate": 4.717741016643119e-05,
+      "loss": 0.9033,
+      "step": 4968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.001099173220473,
+      "learning_rate": 4.717548479743925e-05,
+      "loss": 2.1641,
+      "step": 4969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.922127886518124,
+      "learning_rate": 4.717355881131261e-05,
+      "loss": 1.2578,
+      "step": 4970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7125909253891227,
+      "learning_rate": 4.717163220810484e-05,
+      "loss": 1.3398,
+      "step": 4971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0424761979512835,
+      "learning_rate": 4.716970498786959e-05,
+      "loss": 1.1084,
+      "step": 4972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.2461050485552905,
+      "learning_rate": 4.7167777150660466e-05,
+      "loss": 1.4043,
+      "step": 4973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.959899505397029,
+      "learning_rate": 4.716584869653112e-05,
+      "loss": 1.5547,
+      "step": 4974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.433590785328704,
+      "learning_rate": 4.7163919625535244e-05,
+      "loss": 0.9199,
+      "step": 4975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.836831153545693,
+      "learning_rate": 4.71619899377265e-05,
+      "loss": 0.7959,
+      "step": 4976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5934322116539104,
+      "learning_rate": 4.7160059633158596e-05,
+      "loss": 1.3242,
+      "step": 4977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.552872599308626,
+      "learning_rate": 4.7158128711885264e-05,
+      "loss": 0.8938,
+      "step": 4978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.397590306727685,
+      "learning_rate": 4.715619717396022e-05,
+      "loss": 0.9355,
+      "step": 4979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.860524983060779,
+      "learning_rate": 4.715426501943723e-05,
+      "loss": 1.2969,
+      "step": 4980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9733366218970665,
+      "learning_rate": 4.715233224837007e-05,
+      "loss": 0.9844,
+      "step": 4981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0869114348734628,
+      "learning_rate": 4.7150398860812516e-05,
+      "loss": 0.9736,
+      "step": 4982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.574380057810109,
+      "learning_rate": 4.714846485681837e-05,
+      "loss": 1.584,
+      "step": 4983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9787105302374408,
+      "learning_rate": 4.714653023644148e-05,
+      "loss": 0.3469,
+      "step": 4984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.169207345104655,
+      "learning_rate": 4.714459499973567e-05,
+      "loss": 1.3828,
+      "step": 4985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.36910174156774,
+      "learning_rate": 4.714265914675479e-05,
+      "loss": 1.1475,
+      "step": 4986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4408091482832446,
+      "learning_rate": 4.7140722677552714e-05,
+      "loss": 1.4121,
+      "step": 4987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6840948126124373,
+      "learning_rate": 4.713878559218334e-05,
+      "loss": 1.2783,
+      "step": 4988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.006907841088189,
+      "learning_rate": 4.713684789070059e-05,
+      "loss": 1.1348,
+      "step": 4989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9379022857432067,
+      "learning_rate": 4.713490957315836e-05,
+      "loss": 0.8438,
+      "step": 4990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.413191717897152,
+      "learning_rate": 4.713297063961062e-05,
+      "loss": 0.9189,
+      "step": 4991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6556427040175967,
+      "learning_rate": 4.713103109011131e-05,
+      "loss": 0.8912,
+      "step": 4992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.616918359131738,
+      "learning_rate": 4.7129090924714415e-05,
+      "loss": 1.7656,
+      "step": 4993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.429174534657879,
+      "learning_rate": 4.7127150143473925e-05,
+      "loss": 1.5078,
+      "step": 4994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.474965817865705,
+      "learning_rate": 4.7125208746443864e-05,
+      "loss": 1.5625,
+      "step": 4995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5823493334532746,
+      "learning_rate": 4.712326673367824e-05,
+      "loss": 0.7034,
+      "step": 4996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.559741418437784,
+      "learning_rate": 4.7121324105231116e-05,
+      "loss": 0.9175,
+      "step": 4997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.158540252450371,
+      "learning_rate": 4.7119380861156544e-05,
+      "loss": 1.2188,
+      "step": 4998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0254412179537478,
+      "learning_rate": 4.7117437001508605e-05,
+      "loss": 1.1289,
+      "step": 4999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.122614362890967,
+      "learning_rate": 4.7115492526341406e-05,
+      "loss": 1.3496,
+      "step": 5000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.423540536319843,
+      "learning_rate": 4.711354743570905e-05,
+      "loss": 1.3281,
+      "step": 5001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4563155818187448,
+      "learning_rate": 4.7111601729665664e-05,
+      "loss": 0.718,
+      "step": 5002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.400316455186872,
+      "learning_rate": 4.710965540826542e-05,
+      "loss": 1.252,
+      "step": 5003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.271124902712436,
+      "learning_rate": 4.710770847156245e-05,
+      "loss": 1.2461,
+      "step": 5004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9935283760580518,
+      "learning_rate": 4.710576091961096e-05,
+      "loss": 1.2539,
+      "step": 5005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.033133656827479,
+      "learning_rate": 4.710381275246515e-05,
+      "loss": 2.1328,
+      "step": 5006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.176179758010402,
+      "learning_rate": 4.710186397017922e-05,
+      "loss": 0.8916,
+      "step": 5007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.75939955025626,
+      "learning_rate": 4.709991457280741e-05,
+      "loss": 1.0898,
+      "step": 5008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.604172489398479,
+      "learning_rate": 4.709796456040398e-05,
+      "loss": 1.1211,
+      "step": 5009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.661807641380057,
+      "learning_rate": 4.70960139330232e-05,
+      "loss": 0.8613,
+      "step": 5010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9494789407595805,
+      "learning_rate": 4.709406269071934e-05,
+      "loss": 1.4609,
+      "step": 5011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1389236016918765,
+      "learning_rate": 4.709211083354671e-05,
+      "loss": 1.1865,
+      "step": 5012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8813291539367243,
+      "learning_rate": 4.7090158361559636e-05,
+      "loss": 0.9419,
+      "step": 5013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.127820980734002,
+      "learning_rate": 4.7088205274812435e-05,
+      "loss": 1.6992,
+      "step": 5014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.155143043641306,
+      "learning_rate": 4.7086251573359485e-05,
+      "loss": 2.2295,
+      "step": 5015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.177911881308524,
+      "learning_rate": 4.708429725725514e-05,
+      "loss": 1.0093,
+      "step": 5016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.212129272085645,
+      "learning_rate": 4.708234232655379e-05,
+      "loss": 0.9385,
+      "step": 5017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9143910567961866,
+      "learning_rate": 4.7080386781309844e-05,
+      "loss": 0.6064,
+      "step": 5018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.549365177794938,
+      "learning_rate": 4.707843062157772e-05,
+      "loss": 1.6855,
+      "step": 5019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0840463758907743,
+      "learning_rate": 4.7076473847411864e-05,
+      "loss": 1.293,
+      "step": 5020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.779358590516203,
+      "learning_rate": 4.707451645886674e-05,
+      "loss": 1.1602,
+      "step": 5021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.870619541998779,
+      "learning_rate": 4.707255845599679e-05,
+      "loss": 1.207,
+      "step": 5022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.692114224720787,
+      "learning_rate": 4.707059983885653e-05,
+      "loss": 1.1699,
+      "step": 5023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7700933362198517,
+      "learning_rate": 4.706864060750047e-05,
+      "loss": 0.8633,
+      "step": 5024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.038825724578952,
+      "learning_rate": 4.7066680761983114e-05,
+      "loss": 0.7705,
+      "step": 5025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.5539622416821715,
+      "learning_rate": 4.706472030235902e-05,
+      "loss": 1.2451,
+      "step": 5026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8100447956966126,
+      "learning_rate": 4.706275922868275e-05,
+      "loss": 1.1074,
+      "step": 5027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.329868244024164,
+      "learning_rate": 4.706079754100886e-05,
+      "loss": 1.0195,
+      "step": 5028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1711883557254072,
+      "learning_rate": 4.705883523939196e-05,
+      "loss": 1.3242,
+      "step": 5029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0626893390099093,
+      "learning_rate": 4.705687232388666e-05,
+      "loss": 1.1406,
+      "step": 5030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.509891565066213,
+      "learning_rate": 4.705490879454758e-05,
+      "loss": 0.7974,
+      "step": 5031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.109333593965043,
+      "learning_rate": 4.705294465142936e-05,
+      "loss": 1.2969,
+      "step": 5032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4473890071758473,
+      "learning_rate": 4.7050979894586674e-05,
+      "loss": 0.8301,
+      "step": 5033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9270255690156621,
+      "learning_rate": 4.70490145240742e-05,
+      "loss": 0.4614,
+      "step": 5034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5949597520127132,
+      "learning_rate": 4.7047048539946624e-05,
+      "loss": 0.6084,
+      "step": 5035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.165151371397596,
+      "learning_rate": 4.7045081942258664e-05,
+      "loss": 1.2783,
+      "step": 5036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.532359635852069,
+      "learning_rate": 4.704311473106504e-05,
+      "loss": 2.1406,
+      "step": 5037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.798499524939418,
+      "learning_rate": 4.7041146906420506e-05,
+      "loss": 1.7344,
+      "step": 5038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0156857976689513,
+      "learning_rate": 4.703917846837984e-05,
+      "loss": 1.3906,
+      "step": 5039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.595619954661141,
+      "learning_rate": 4.70372094169978e-05,
+      "loss": 1.4448,
+      "step": 5040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.736640732611727,
+      "learning_rate": 4.7035239752329195e-05,
+      "loss": 1.0742,
+      "step": 5041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0892381022225113,
+      "learning_rate": 4.7033269474428843e-05,
+      "loss": 0.7588,
+      "step": 5042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.630355186265594,
+      "learning_rate": 4.703129858335156e-05,
+      "loss": 1.6348,
+      "step": 5043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6195921186520055,
+      "learning_rate": 4.702932707915222e-05,
+      "loss": 1.3281,
+      "step": 5044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0259278758064476,
+      "learning_rate": 4.702735496188566e-05,
+      "loss": 0.8818,
+      "step": 5045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.788039429555639,
+      "learning_rate": 4.702538223160679e-05,
+      "loss": 0.7622,
+      "step": 5046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.511000376241474,
+      "learning_rate": 4.70234088883705e-05,
+      "loss": 0.6309,
+      "step": 5047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.668729556213997,
+      "learning_rate": 4.70214349322317e-05,
+      "loss": 1.1895,
+      "step": 5048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.105783660084764,
+      "learning_rate": 4.701946036324534e-05,
+      "loss": 1.0938,
+      "step": 5049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.874460630656415,
+      "learning_rate": 4.701748518146635e-05,
+      "loss": 0.7783,
+      "step": 5050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.776692820562545,
+      "learning_rate": 4.701550938694972e-05,
+      "loss": 1.1396,
+      "step": 5051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.304376510069197,
+      "learning_rate": 4.701353297975042e-05,
+      "loss": 1.166,
+      "step": 5052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.709330673471156,
+      "learning_rate": 4.701155595992346e-05,
+      "loss": 0.8342,
+      "step": 5053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.847103410131236,
+      "learning_rate": 4.700957832752386e-05,
+      "loss": 1.0188,
+      "step": 5054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7310047414899836,
+      "learning_rate": 4.700760008260666e-05,
+      "loss": 1.1567,
+      "step": 5055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6687064277102044,
+      "learning_rate": 4.70056212252269e-05,
+      "loss": 0.5278,
+      "step": 5056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.281853996815584,
+      "learning_rate": 4.700364175543967e-05,
+      "loss": 1.543,
+      "step": 5057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.94689219076375,
+      "learning_rate": 4.700166167330003e-05,
+      "loss": 1.2422,
+      "step": 5058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.475024591188893,
+      "learning_rate": 4.699968097886312e-05,
+      "loss": 0.7842,
+      "step": 5059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.248029111261933,
+      "learning_rate": 4.699769967218404e-05,
+      "loss": 1.3955,
+      "step": 5060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.261164832928454,
+      "learning_rate": 4.6995717753317935e-05,
+      "loss": 1.0146,
+      "step": 5061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.948172125835321,
+      "learning_rate": 4.699373522231996e-05,
+      "loss": 2.1758,
+      "step": 5062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.436134344536548,
+      "learning_rate": 4.699175207924528e-05,
+      "loss": 1.0767,
+      "step": 5063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3150298336752173,
+      "learning_rate": 4.6989768324149095e-05,
+      "loss": 0.8564,
+      "step": 5064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.108244529432041,
+      "learning_rate": 4.698778395708661e-05,
+      "loss": 0.6348,
+      "step": 5065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9749860655342437,
+      "learning_rate": 4.698579897811305e-05,
+      "loss": 0.9092,
+      "step": 5066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.942083428463303,
+      "learning_rate": 4.6983813387283655e-05,
+      "loss": 1.7207,
+      "step": 5067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.546911528184949,
+      "learning_rate": 4.698182718465368e-05,
+      "loss": 0.9231,
+      "step": 5068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6026117281824983,
+      "learning_rate": 4.69798403702784e-05,
+      "loss": 0.9092,
+      "step": 5069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3534175123060384,
+      "learning_rate": 4.6977852944213124e-05,
+      "loss": 1.6816,
+      "step": 5070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3816953580819558,
+      "learning_rate": 4.697586490651314e-05,
+      "loss": 1.3145,
+      "step": 5071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5390790724208423,
+      "learning_rate": 4.6973876257233784e-05,
+      "loss": 0.7241,
+      "step": 5072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.745400850975081,
+      "learning_rate": 4.6971886996430394e-05,
+      "loss": 1.248,
+      "step": 5073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.089786985396036,
+      "learning_rate": 4.696989712415834e-05,
+      "loss": 1.2773,
+      "step": 5074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.461885275135952,
+      "learning_rate": 4.696790664047298e-05,
+      "loss": 0.9043,
+      "step": 5075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.065668127310533,
+      "learning_rate": 4.696591554542973e-05,
+      "loss": 1.0332,
+      "step": 5076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.351889349949163,
+      "learning_rate": 4.696392383908399e-05,
+      "loss": 1.5273,
+      "step": 5077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7570172857296806,
+      "learning_rate": 4.6961931521491196e-05,
+      "loss": 0.7461,
+      "step": 5078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0213895263777197,
+      "learning_rate": 4.6959938592706786e-05,
+      "loss": 1.793,
+      "step": 5079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.187958294950105,
+      "learning_rate": 4.695794505278622e-05,
+      "loss": 1.8145,
+      "step": 5080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.324224704645291,
+      "learning_rate": 4.695595090178499e-05,
+      "loss": 0.8252,
+      "step": 5081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.827650719714261,
+      "learning_rate": 4.6953956139758584e-05,
+      "loss": 1.2139,
+      "step": 5082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.284802502831027,
+      "learning_rate": 4.695196076676251e-05,
+      "loss": 1.3945,
+      "step": 5083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.42589761697539,
+      "learning_rate": 4.694996478285231e-05,
+      "loss": 1.082,
+      "step": 5084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9105455381405467,
+      "learning_rate": 4.694796818808352e-05,
+      "loss": 1.0352,
+      "step": 5085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1686800194711835,
+      "learning_rate": 4.6945970982511726e-05,
+      "loss": 1.1992,
+      "step": 5086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3536645359400095,
+      "learning_rate": 4.694397316619248e-05,
+      "loss": 0.9434,
+      "step": 5087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.312482679147548,
+      "learning_rate": 4.6941974739181395e-05,
+      "loss": 1.4648,
+      "step": 5088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.811241383609233,
+      "learning_rate": 4.6939975701534086e-05,
+      "loss": 1.1562,
+      "step": 5089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.979789514038317,
+      "learning_rate": 4.693797605330619e-05,
+      "loss": 0.9805,
+      "step": 5090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.48841650039243,
+      "learning_rate": 4.6935975794553345e-05,
+      "loss": 1.2754,
+      "step": 5091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6134031350032703,
+      "learning_rate": 4.693397492533123e-05,
+      "loss": 0.8223,
+      "step": 5092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.480701922828365,
+      "learning_rate": 4.6931973445695506e-05,
+      "loss": 1.5693,
+      "step": 5093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0767506276002408,
+      "learning_rate": 4.6929971355701905e-05,
+      "loss": 1.4707,
+      "step": 5094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.547449526899807,
+      "learning_rate": 4.692796865540613e-05,
+      "loss": 1.6035,
+      "step": 5095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.66942549561969,
+      "learning_rate": 4.692596534486391e-05,
+      "loss": 1.7188,
+      "step": 5096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9365653487907823,
+      "learning_rate": 4.6923961424131e-05,
+      "loss": 1.2412,
+      "step": 5097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.211542372706761,
+      "learning_rate": 4.692195689326316e-05,
+      "loss": 1.1064,
+      "step": 5098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3728683341721974,
+      "learning_rate": 4.6919951752316196e-05,
+      "loss": 1.5791,
+      "step": 5099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.684384780528466,
+      "learning_rate": 4.69179460013459e-05,
+      "loss": 1.5195,
+      "step": 5100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5851036179396405,
+      "learning_rate": 4.691593964040808e-05,
+      "loss": 1.0508,
+      "step": 5101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.116735656495052,
+      "learning_rate": 4.691393266955859e-05,
+      "loss": 1.3086,
+      "step": 5102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.226098263178299,
+      "learning_rate": 4.691192508885327e-05,
+      "loss": 0.3691,
+      "step": 5103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0377268074662864,
+      "learning_rate": 4.6909916898348e-05,
+      "loss": 1.6875,
+      "step": 5104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8087363504954563,
+      "learning_rate": 4.690790809809866e-05,
+      "loss": 1.2598,
+      "step": 5105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.595694399116494,
+      "learning_rate": 4.690589868816115e-05,
+      "loss": 1.0135,
+      "step": 5106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.12769875536007,
+      "learning_rate": 4.6903888668591414e-05,
+      "loss": 0.915,
+      "step": 5107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.907002498681736,
+      "learning_rate": 4.6901878039445354e-05,
+      "loss": 2.0078,
+      "step": 5108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.733528095456298,
+      "learning_rate": 4.689986680077896e-05,
+      "loss": 1.5176,
+      "step": 5109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.583153801494032,
+      "learning_rate": 4.689785495264818e-05,
+      "loss": 1.5166,
+      "step": 5110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4221570250673783,
+      "learning_rate": 4.6895842495109014e-05,
+      "loss": 1.0293,
+      "step": 5111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8252634145082194,
+      "learning_rate": 4.689382942821746e-05,
+      "loss": 0.2886,
+      "step": 5112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.368307320892419,
+      "learning_rate": 4.689181575202955e-05,
+      "loss": 1.4648,
+      "step": 5113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.456729854225698,
+      "learning_rate": 4.688980146660132e-05,
+      "loss": 1.019,
+      "step": 5114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2126163446755003,
+      "learning_rate": 4.6887786571988824e-05,
+      "loss": 0.6836,
+      "step": 5115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.555020406867835,
+      "learning_rate": 4.688577106824814e-05,
+      "loss": 1.0879,
+      "step": 5116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.665586332194995,
+      "learning_rate": 4.688375495543536e-05,
+      "loss": 1.3027,
+      "step": 5117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.686176367481655,
+      "learning_rate": 4.6881738233606584e-05,
+      "loss": 1.1777,
+      "step": 5118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.665892377832836,
+      "learning_rate": 4.687972090281794e-05,
+      "loss": 0.5449,
+      "step": 5119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.439270072466452,
+      "learning_rate": 4.6877702963125574e-05,
+      "loss": 1.25,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.871419794588297,
+      "learning_rate": 4.687568441458564e-05,
+      "loss": 0.9951,
+      "step": 5121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4742058813990364,
+      "learning_rate": 4.687366525725432e-05,
+      "loss": 0.9189,
+      "step": 5122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2586778548038713,
+      "learning_rate": 4.687164549118779e-05,
+      "loss": 1.623,
+      "step": 5123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.2610301450044465,
+      "learning_rate": 4.6869625116442274e-05,
+      "loss": 1.2285,
+      "step": 5124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9236825026183835,
+      "learning_rate": 4.6867604133073995e-05,
+      "loss": 0.7935,
+      "step": 5125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.690703917968925,
+      "learning_rate": 4.6865582541139194e-05,
+      "loss": 0.5432,
+      "step": 5126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.21529732286633,
+      "learning_rate": 4.6863560340694134e-05,
+      "loss": 1.0159,
+      "step": 5127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.661083894332492,
+      "learning_rate": 4.6861537531795094e-05,
+      "loss": 1.1387,
+      "step": 5128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7150008560938037,
+      "learning_rate": 4.685951411449836e-05,
+      "loss": 0.8594,
+      "step": 5129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.305290014098613,
+      "learning_rate": 4.6857490088860244e-05,
+      "loss": 0.8669,
+      "step": 5130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.139720196480389,
+      "learning_rate": 4.685546545493709e-05,
+      "loss": 1.082,
+      "step": 5131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.5977900379456464,
+      "learning_rate": 4.685344021278522e-05,
+      "loss": 1.0518,
+      "step": 5132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1679965957565512,
+      "learning_rate": 4.6851414362461e-05,
+      "loss": 0.9971,
+      "step": 5133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8788240084040146,
+      "learning_rate": 4.684938790402082e-05,
+      "loss": 1.0029,
+      "step": 5134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.9519012576043435,
+      "learning_rate": 4.684736083752107e-05,
+      "loss": 1.2598,
+      "step": 5135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.193368189396431,
+      "learning_rate": 4.684533316301816e-05,
+      "loss": 1.2295,
+      "step": 5136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.856683301010059,
+      "learning_rate": 4.6843304880568523e-05,
+      "loss": 1.2852,
+      "step": 5137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0263816830687054,
+      "learning_rate": 4.6841275990228603e-05,
+      "loss": 1.4531,
+      "step": 5138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5166838683179216,
+      "learning_rate": 4.6839246492054866e-05,
+      "loss": 0.8818,
+      "step": 5139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.412027136755983,
+      "learning_rate": 4.683721638610379e-05,
+      "loss": 0.6309,
+      "step": 5140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.054611322179934,
+      "learning_rate": 4.683518567243187e-05,
+      "loss": 1.2617,
+      "step": 5141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6645000020215246,
+      "learning_rate": 4.683315435109562e-05,
+      "loss": 1.3359,
+      "step": 5142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.618502613846452,
+      "learning_rate": 4.683112242215157e-05,
+      "loss": 0.8848,
+      "step": 5143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.051315150926154,
+      "learning_rate": 4.682908988565627e-05,
+      "loss": 0.7695,
+      "step": 5144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1532393736559925,
+      "learning_rate": 4.6827056741666295e-05,
+      "loss": 1.1895,
+      "step": 5145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.702588970535885,
+      "learning_rate": 4.682502299023821e-05,
+      "loss": 1.6455,
+      "step": 5146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.579935282287448,
+      "learning_rate": 4.682298863142862e-05,
+      "loss": 1.2041,
+      "step": 5147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.834805758653755,
+      "learning_rate": 4.682095366529415e-05,
+      "loss": 1.0215,
+      "step": 5148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.93241974483575,
+      "learning_rate": 4.6818918091891415e-05,
+      "loss": 1.6641,
+      "step": 5149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.175393190051444,
+      "learning_rate": 4.6816881911277066e-05,
+      "loss": 1.4023,
+      "step": 5150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2734867890248682,
+      "learning_rate": 4.6814845123507786e-05,
+      "loss": 0.8281,
+      "step": 5151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9478384260142967,
+      "learning_rate": 4.681280772864024e-05,
+      "loss": 0.9209,
+      "step": 5152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0309490565716732,
+      "learning_rate": 4.681076972673114e-05,
+      "loss": 1.2598,
+      "step": 5153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4384170176638102,
+      "learning_rate": 4.680873111783719e-05,
+      "loss": 0.9521,
+      "step": 5154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.211899216577882,
+      "learning_rate": 4.680669190201514e-05,
+      "loss": 0.6382,
+      "step": 5155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0379823466548364,
+      "learning_rate": 4.6804652079321726e-05,
+      "loss": 0.5967,
+      "step": 5156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8982596535628438,
+      "learning_rate": 4.680261164981372e-05,
+      "loss": 0.874,
+      "step": 5157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.287827611090938,
+      "learning_rate": 4.6800570613547914e-05,
+      "loss": 1.1738,
+      "step": 5158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.412764258752139,
+      "learning_rate": 4.6798528970581104e-05,
+      "loss": 0.9707,
+      "step": 5159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.048675560521141,
+      "learning_rate": 4.67964867209701e-05,
+      "loss": 1.0308,
+      "step": 5160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.65505829052057,
+      "learning_rate": 4.679444386477174e-05,
+      "loss": 0.9697,
+      "step": 5161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7385221939805646,
+      "learning_rate": 4.679240040204289e-05,
+      "loss": 0.666,
+      "step": 5162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.402080984349993,
+      "learning_rate": 4.679035633284041e-05,
+      "loss": 0.9971,
+      "step": 5163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.565880972849329,
+      "learning_rate": 4.6788311657221176e-05,
+      "loss": 0.7549,
+      "step": 5164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.138556289677012,
+      "learning_rate": 4.6786266375242104e-05,
+      "loss": 1.4648,
+      "step": 5165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.486345289580137,
+      "learning_rate": 4.678422048696011e-05,
+      "loss": 1.0889,
+      "step": 5166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.237425883947499,
+      "learning_rate": 4.6782173992432116e-05,
+      "loss": 1.2188,
+      "step": 5167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.654358762759241,
+      "learning_rate": 4.6780126891715095e-05,
+      "loss": 0.6578,
+      "step": 5168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.655834000200905,
+      "learning_rate": 4.6778079184866006e-05,
+      "loss": 1.1523,
+      "step": 5169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4081924326861173,
+      "learning_rate": 4.6776030871941844e-05,
+      "loss": 1.2383,
+      "step": 5170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2752294620178337,
+      "learning_rate": 4.677398195299961e-05,
+      "loss": 0.9814,
+      "step": 5171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.590836324689023,
+      "learning_rate": 4.677193242809632e-05,
+      "loss": 1.2852,
+      "step": 5172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.555855486955911,
+      "learning_rate": 4.6769882297289006e-05,
+      "loss": 0.5664,
+      "step": 5173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3294835429600487,
+      "learning_rate": 4.6767831560634744e-05,
+      "loss": 1.3477,
+      "step": 5174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.941856762896846,
+      "learning_rate": 4.676578021819058e-05,
+      "loss": 1.4258,
+      "step": 5175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8943631395120484,
+      "learning_rate": 4.6763728270013616e-05,
+      "loss": 0.5479,
+      "step": 5176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.338208373744276,
+      "learning_rate": 4.676167571616096e-05,
+      "loss": 1.3457,
+      "step": 5177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9395621346065965,
+      "learning_rate": 4.6759622556689716e-05,
+      "loss": 1.0459,
+      "step": 5178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3783180033469735,
+      "learning_rate": 4.675756879165705e-05,
+      "loss": 1.3672,
+      "step": 5179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.858487342075021,
+      "learning_rate": 4.67555144211201e-05,
+      "loss": 0.6924,
+      "step": 5180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.633542828334925,
+      "learning_rate": 4.675345944513603e-05,
+      "loss": 0.3569,
+      "step": 5181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.854615466248308,
+      "learning_rate": 4.675140386376205e-05,
+      "loss": 1.3672,
+      "step": 5182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.8073076006671815,
+      "learning_rate": 4.674934767705534e-05,
+      "loss": 1.6133,
+      "step": 5183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9578124290663785,
+      "learning_rate": 4.674729088507315e-05,
+      "loss": 1.125,
+      "step": 5184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3042513159803697,
+      "learning_rate": 4.674523348787271e-05,
+      "loss": 0.8447,
+      "step": 5185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.122571288001968,
+      "learning_rate": 4.674317548551127e-05,
+      "loss": 0.8389,
+      "step": 5186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.873953255546004,
+      "learning_rate": 4.674111687804611e-05,
+      "loss": 1.4014,
+      "step": 5187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9163261263084728,
+      "learning_rate": 4.673905766553451e-05,
+      "loss": 0.731,
+      "step": 5188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.608199911087905,
+      "learning_rate": 4.673699784803379e-05,
+      "loss": 1.7656,
+      "step": 5189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.504365242416706,
+      "learning_rate": 4.673493742560128e-05,
+      "loss": 0.9727,
+      "step": 5190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6672862843040535,
+      "learning_rate": 4.6732876398294294e-05,
+      "loss": 1.2129,
+      "step": 5191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2680166315010526,
+      "learning_rate": 4.673081476617022e-05,
+      "loss": 1.2031,
+      "step": 5192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.68185159474323,
+      "learning_rate": 4.67287525292864e-05,
+      "loss": 0.938,
+      "step": 5193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.058318884570286,
+      "learning_rate": 4.6726689687700254e-05,
+      "loss": 0.9258,
+      "step": 5194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.577380541041354,
+      "learning_rate": 4.6724626241469186e-05,
+      "loss": 0.7471,
+      "step": 5195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.463374903943433,
+      "learning_rate": 4.67225621906506e-05,
+      "loss": 0.876,
+      "step": 5196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4452936823794227,
+      "learning_rate": 4.672049753530195e-05,
+      "loss": 0.9648,
+      "step": 5197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.552671686409675,
+      "learning_rate": 4.6718432275480706e-05,
+      "loss": 0.9341,
+      "step": 5198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.606881312239695,
+      "learning_rate": 4.671636641124433e-05,
+      "loss": 2.2188,
+      "step": 5199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.613387534769933,
+      "learning_rate": 4.671429994265032e-05,
+      "loss": 1.165,
+      "step": 5200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.60497154513174,
+      "learning_rate": 4.6712232869756175e-05,
+      "loss": 0.895,
+      "step": 5201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.353144218295421,
+      "learning_rate": 4.6710165192619436e-05,
+      "loss": 1.2324,
+      "step": 5202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.923752644158386,
+      "learning_rate": 4.670809691129764e-05,
+      "loss": 0.9189,
+      "step": 5203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.08174925199684,
+      "learning_rate": 4.6706028025848335e-05,
+      "loss": 0.7637,
+      "step": 5204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.027298071732003,
+      "learning_rate": 4.670395853632911e-05,
+      "loss": 0.8115,
+      "step": 5205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.465239526966072,
+      "learning_rate": 4.670188844279756e-05,
+      "loss": 1.4189,
+      "step": 5206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.400174978418272,
+      "learning_rate": 4.669981774531128e-05,
+      "loss": 0.8779,
+      "step": 5207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7766847324890334,
+      "learning_rate": 4.669774644392791e-05,
+      "loss": 1.6055,
+      "step": 5208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.763722124260907,
+      "learning_rate": 4.669567453870509e-05,
+      "loss": 0.5469,
+      "step": 5209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8052683549406594,
+      "learning_rate": 4.669360202970049e-05,
+      "loss": 0.8604,
+      "step": 5210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3058075866645398,
+      "learning_rate": 4.669152891697176e-05,
+      "loss": 0.6935,
+      "step": 5211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1780202182537156,
+      "learning_rate": 4.668945520057662e-05,
+      "loss": 1.4355,
+      "step": 5212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8059740339186225,
+      "learning_rate": 4.668738088057277e-05,
+      "loss": 1.2236,
+      "step": 5213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.099628906159158,
+      "learning_rate": 4.6685305957017945e-05,
+      "loss": 0.6646,
+      "step": 5214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.368147274197188,
+      "learning_rate": 4.668323042996987e-05,
+      "loss": 0.959,
+      "step": 5215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.83261460182689,
+      "learning_rate": 4.668115429948633e-05,
+      "loss": 1.0898,
+      "step": 5216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.150107790614864,
+      "learning_rate": 4.6679077565625084e-05,
+      "loss": 1.7871,
+      "step": 5217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.387193622317862,
+      "learning_rate": 4.6677000228443946e-05,
+      "loss": 0.9717,
+      "step": 5218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4149747933282635,
+      "learning_rate": 4.667492228800071e-05,
+      "loss": 0.8447,
+      "step": 5219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.594685149253684,
+      "learning_rate": 4.667284374435321e-05,
+      "loss": 0.6908,
+      "step": 5220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.070556320332361,
+      "learning_rate": 4.667076459755929e-05,
+      "loss": 0.6807,
+      "step": 5221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5873230392467257,
+      "learning_rate": 4.666868484767682e-05,
+      "loss": 0.9027,
+      "step": 5222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.988293904711859,
+      "learning_rate": 4.666660449476366e-05,
+      "loss": 1.0771,
+      "step": 5223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4839382147545495,
+      "learning_rate": 4.666452353887773e-05,
+      "loss": 0.8899,
+      "step": 5224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.480375652107071,
+      "learning_rate": 4.666244198007693e-05,
+      "loss": 1.2656,
+      "step": 5225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.831018780423657,
+      "learning_rate": 4.666035981841918e-05,
+      "loss": 1.5762,
+      "step": 5226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6091342862271705,
+      "learning_rate": 4.665827705396244e-05,
+      "loss": 1.3359,
+      "step": 5227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.023794206024696,
+      "learning_rate": 4.665619368676466e-05,
+      "loss": 1.418,
+      "step": 5228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.207613893242539,
+      "learning_rate": 4.665410971688383e-05,
+      "loss": 1.3574,
+      "step": 5229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4202415713383516,
+      "learning_rate": 4.665202514437794e-05,
+      "loss": 0.9287,
+      "step": 5230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6177566137287673,
+      "learning_rate": 4.6649939969305015e-05,
+      "loss": 0.8569,
+      "step": 5231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5045406115242312,
+      "learning_rate": 4.664785419172306e-05,
+      "loss": 1.2373,
+      "step": 5232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.40202743804913,
+      "learning_rate": 4.664576781169014e-05,
+      "loss": 1.2383,
+      "step": 5233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.784074186441262,
+      "learning_rate": 4.6643680829264324e-05,
+      "loss": 1.25,
+      "step": 5234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.327905222100654,
+      "learning_rate": 4.664159324450367e-05,
+      "loss": 0.7249,
+      "step": 5235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.123990439115659,
+      "learning_rate": 4.663950505746629e-05,
+      "loss": 1.6855,
+      "step": 5236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.760368675251572,
+      "learning_rate": 4.663741626821029e-05,
+      "loss": 0.793,
+      "step": 5237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.299203186597189,
+      "learning_rate": 4.663532687679381e-05,
+      "loss": 1.2109,
+      "step": 5238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.156272556352808,
+      "learning_rate": 4.663323688327499e-05,
+      "loss": 1.1162,
+      "step": 5239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9901814252493955,
+      "learning_rate": 4.6631146287711994e-05,
+      "loss": 0.9585,
+      "step": 5240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.046614535700179,
+      "learning_rate": 4.6629055090163e-05,
+      "loss": 0.8252,
+      "step": 5241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.304632404955995,
+      "learning_rate": 4.662696329068621e-05,
+      "loss": 1.0415,
+      "step": 5242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4906394309323594,
+      "learning_rate": 4.662487088933984e-05,
+      "loss": 1.3008,
+      "step": 5243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.887112169415115,
+      "learning_rate": 4.6622777886182106e-05,
+      "loss": 1.1377,
+      "step": 5244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.223860520415282,
+      "learning_rate": 4.662068428127128e-05,
+      "loss": 1.0283,
+      "step": 5245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4828908628251827,
+      "learning_rate": 4.6618590074665596e-05,
+      "loss": 0.7026,
+      "step": 5246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.941818666398777,
+      "learning_rate": 4.661649526642336e-05,
+      "loss": 1.1602,
+      "step": 5247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.3285628004406265,
+      "learning_rate": 4.6614399856602864e-05,
+      "loss": 0.9316,
+      "step": 5248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.919336084078776,
+      "learning_rate": 4.6612303845262414e-05,
+      "loss": 0.9873,
+      "step": 5249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1794199352046144,
+      "learning_rate": 4.661020723246035e-05,
+      "loss": 0.5381,
+      "step": 5250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9672494408769827,
+      "learning_rate": 4.6608110018255016e-05,
+      "loss": 0.8799,
+      "step": 5251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.795229123846087,
+      "learning_rate": 4.660601220270477e-05,
+      "loss": 0.4225,
+      "step": 5252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6259406992774657,
+      "learning_rate": 4.6603913785868005e-05,
+      "loss": 1.25,
+      "step": 5253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.712801421685213,
+      "learning_rate": 4.6601814767803115e-05,
+      "loss": 1.2852,
+      "step": 5254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.518683565369743,
+      "learning_rate": 4.659971514856851e-05,
+      "loss": 1.0605,
+      "step": 5255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.290836043977773,
+      "learning_rate": 4.6597614928222635e-05,
+      "loss": 1.0107,
+      "step": 5256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5961628916438833,
+      "learning_rate": 4.659551410682392e-05,
+      "loss": 0.6831,
+      "step": 5257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.828751973999791,
+      "learning_rate": 4.6593412684430844e-05,
+      "loss": 1.4434,
+      "step": 5258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8773540313313295,
+      "learning_rate": 4.659131066110188e-05,
+      "loss": 1.0527,
+      "step": 5259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.674795405352963,
+      "learning_rate": 4.6589208036895535e-05,
+      "loss": 0.4761,
+      "step": 5260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5260645510195276,
+      "learning_rate": 4.658710481187032e-05,
+      "loss": 0.9013,
+      "step": 5261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.027345396419731,
+      "learning_rate": 4.658500098608476e-05,
+      "loss": 1.7461,
+      "step": 5262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2729458096946775,
+      "learning_rate": 4.658289655959741e-05,
+      "loss": 1.0674,
+      "step": 5263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.307143898899796,
+      "learning_rate": 4.658079153246684e-05,
+      "loss": 1.1523,
+      "step": 5264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.477726515376053,
+      "learning_rate": 4.6578685904751616e-05,
+      "loss": 0.8135,
+      "step": 5265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.1214762577963615,
+      "learning_rate": 4.6576579676510354e-05,
+      "loss": 0.9385,
+      "step": 5266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.097403257576705,
+      "learning_rate": 4.6574472847801677e-05,
+      "loss": 1.1279,
+      "step": 5267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.576427293772843,
+      "learning_rate": 4.657236541868419e-05,
+      "loss": 0.7236,
+      "step": 5268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.351315000569485,
+      "learning_rate": 4.657025738921655e-05,
+      "loss": 1.4062,
+      "step": 5269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5725572737099665,
+      "learning_rate": 4.656814875945744e-05,
+      "loss": 1.5195,
+      "step": 5270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.340985875848569,
+      "learning_rate": 4.656603952946553e-05,
+      "loss": 0.6533,
+      "step": 5271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.601168728897503,
+      "learning_rate": 4.6563929699299514e-05,
+      "loss": 1.0469,
+      "step": 5272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6168302355458692,
+      "learning_rate": 4.6561819269018115e-05,
+      "loss": 1.5723,
+      "step": 5273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9403734741362215,
+      "learning_rate": 4.6559708238680064e-05,
+      "loss": 1.1836,
+      "step": 5274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.561751279540053,
+      "learning_rate": 4.655759660834411e-05,
+      "loss": 0.9336,
+      "step": 5275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.9883410899039005,
+      "learning_rate": 4.655548437806902e-05,
+      "loss": 1.3564,
+      "step": 5276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6431275196293447,
+      "learning_rate": 4.655337154791357e-05,
+      "loss": 1.5273,
+      "step": 5277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5458421543632395,
+      "learning_rate": 4.655125811793657e-05,
+      "loss": 1.2812,
+      "step": 5278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.308525071518362,
+      "learning_rate": 4.654914408819683e-05,
+      "loss": 1.0768,
+      "step": 5279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.496468960030805,
+      "learning_rate": 4.654702945875318e-05,
+      "loss": 1.3379,
+      "step": 5280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.055256463821505,
+      "learning_rate": 4.654491422966447e-05,
+      "loss": 1.0068,
+      "step": 5281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.324779773904497,
+      "learning_rate": 4.654279840098957e-05,
+      "loss": 1.5498,
+      "step": 5282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.788582416737457,
+      "learning_rate": 4.6540681972787356e-05,
+      "loss": 1.209,
+      "step": 5283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.751664121143877,
+      "learning_rate": 4.653856494511674e-05,
+      "loss": 1.5156,
+      "step": 5284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.6528789427025,
+      "learning_rate": 4.6536447318036625e-05,
+      "loss": 0.9854,
+      "step": 5285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6483888806650735,
+      "learning_rate": 4.653432909160595e-05,
+      "loss": 1.2451,
+      "step": 5286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.4163450683589325,
+      "learning_rate": 4.6532210265883665e-05,
+      "loss": 0.7073,
+      "step": 5287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1932796963140446,
+      "learning_rate": 4.653009084092873e-05,
+      "loss": 0.4155,
+      "step": 5288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7748531742468585,
+      "learning_rate": 4.652797081680013e-05,
+      "loss": 1.4766,
+      "step": 5289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8063502062816514,
+      "learning_rate": 4.652585019355687e-05,
+      "loss": 1.0,
+      "step": 5290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2775691396894353,
+      "learning_rate": 4.6523728971257966e-05,
+      "loss": 0.9124,
+      "step": 5291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.416536316699581,
+      "learning_rate": 4.6521607149962444e-05,
+      "loss": 0.6825,
+      "step": 5292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3694732059580095,
+      "learning_rate": 4.651948472972936e-05,
+      "loss": 1.4707,
+      "step": 5293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.641295501505193,
+      "learning_rate": 4.6517361710617765e-05,
+      "loss": 0.792,
+      "step": 5294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.345115311578049,
+      "learning_rate": 4.6515238092686765e-05,
+      "loss": 1.252,
+      "step": 5295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.058834580259414,
+      "learning_rate": 4.651311387599544e-05,
+      "loss": 1.4902,
+      "step": 5296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0407038254285936,
+      "learning_rate": 4.6510989060602916e-05,
+      "loss": 1.0195,
+      "step": 5297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0967709982891414,
+      "learning_rate": 4.6508863646568326e-05,
+      "loss": 0.9961,
+      "step": 5298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.199772658305849,
+      "learning_rate": 4.650673763395081e-05,
+      "loss": 0.9559,
+      "step": 5299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5305186969554025,
+      "learning_rate": 4.650461102280955e-05,
+      "loss": 1.1201,
+      "step": 5300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.843905869109296,
+      "learning_rate": 4.650248381320372e-05,
+      "loss": 1.3296,
+      "step": 5301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.943840326757856,
+      "learning_rate": 4.6500356005192514e-05,
+      "loss": 0.9844,
+      "step": 5302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2678459068509174,
+      "learning_rate": 4.649822759883515e-05,
+      "loss": 0.9189,
+      "step": 5303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7906513738104626,
+      "learning_rate": 4.649609859419087e-05,
+      "loss": 0.8831,
+      "step": 5304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8115467575394013,
+      "learning_rate": 4.649396899131892e-05,
+      "loss": 1.1055,
+      "step": 5305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5692669907427796,
+      "learning_rate": 4.649183879027857e-05,
+      "loss": 0.8984,
+      "step": 5306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.590016078825423,
+      "learning_rate": 4.648970799112907e-05,
+      "loss": 0.9009,
+      "step": 5307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2974641824878157,
+      "learning_rate": 4.6487576593929765e-05,
+      "loss": 0.8242,
+      "step": 5308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9984020379160876,
+      "learning_rate": 4.6485444598739944e-05,
+      "loss": 0.9424,
+      "step": 5309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.919382914761783,
+      "learning_rate": 4.648331200561895e-05,
+      "loss": 1.748,
+      "step": 5310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0912110670748536,
+      "learning_rate": 4.648117881462612e-05,
+      "loss": 0.4692,
+      "step": 5311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.746303525456597,
+      "learning_rate": 4.647904502582085e-05,
+      "loss": 1.3262,
+      "step": 5312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.72554715637581,
+      "learning_rate": 4.6476910639262483e-05,
+      "loss": 1.6992,
+      "step": 5313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5553269825408678,
+      "learning_rate": 4.647477565501044e-05,
+      "loss": 1.1289,
+      "step": 5314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.48911787547399,
+      "learning_rate": 4.6472640073124126e-05,
+      "loss": 1.2207,
+      "step": 5315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7457201376598346,
+      "learning_rate": 4.6470503893662995e-05,
+      "loss": 1.0645,
+      "step": 5316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1224048700120757,
+      "learning_rate": 4.646836711668647e-05,
+      "loss": 0.6758,
+      "step": 5317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0832113465243385,
+      "learning_rate": 4.646622974225403e-05,
+      "loss": 0.8237,
+      "step": 5318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8826225126824743,
+      "learning_rate": 4.6464091770425156e-05,
+      "loss": 0.9097,
+      "step": 5319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4222113735748123,
+      "learning_rate": 4.646195320125935e-05,
+      "loss": 0.4619,
+      "step": 5320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.128533167345552,
+      "learning_rate": 4.645981403481612e-05,
+      "loss": 2.2383,
+      "step": 5321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7944757837299776,
+      "learning_rate": 4.6457674271155e-05,
+      "loss": 1.3398,
+      "step": 5322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.078230542773402,
+      "learning_rate": 4.645553391033555e-05,
+      "loss": 1.1846,
+      "step": 5323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.041222586922585,
+      "learning_rate": 4.645339295241731e-05,
+      "loss": 0.7725,
+      "step": 5324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.628292046210421,
+      "learning_rate": 4.645125139745989e-05,
+      "loss": 0.7129,
+      "step": 5325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.855702175068023,
+      "learning_rate": 4.644910924552287e-05,
+      "loss": 1.252,
+      "step": 5326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.798808060865423,
+      "learning_rate": 4.644696649666588e-05,
+      "loss": 1.0664,
+      "step": 5327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.197761623008039,
+      "learning_rate": 4.644482315094853e-05,
+      "loss": 0.9707,
+      "step": 5328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.926351781924679,
+      "learning_rate": 4.644267920843049e-05,
+      "loss": 1.627,
+      "step": 5329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6613144368805883,
+      "learning_rate": 4.644053466917142e-05,
+      "loss": 0.7318,
+      "step": 5330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3352073115065446,
+      "learning_rate": 4.6438389533230993e-05,
+      "loss": 1.2227,
+      "step": 5331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.880309302622198,
+      "learning_rate": 4.643624380066891e-05,
+      "loss": 1.2656,
+      "step": 5332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9565817971361352,
+      "learning_rate": 4.6434097471544904e-05,
+      "loss": 0.9141,
+      "step": 5333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.700349121127963,
+      "learning_rate": 4.643195054591867e-05,
+      "loss": 0.7549,
+      "step": 5334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.123281178577603,
+      "learning_rate": 4.642980302385e-05,
+      "loss": 1.0762,
+      "step": 5335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.419430194009342,
+      "learning_rate": 4.6427654905398623e-05,
+      "loss": 0.915,
+      "step": 5336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.720765139216621,
+      "learning_rate": 4.6425506190624325e-05,
+      "loss": 0.6001,
+      "step": 5337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.734822263922572,
+      "learning_rate": 4.6423356879586924e-05,
+      "loss": 0.606,
+      "step": 5338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.737058577694165,
+      "learning_rate": 4.642120697234622e-05,
+      "loss": 1.9238,
+      "step": 5339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1199159952830406,
+      "learning_rate": 4.6419056468962045e-05,
+      "loss": 1.4121,
+      "step": 5340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.418059843087191,
+      "learning_rate": 4.641690536949425e-05,
+      "loss": 1.1289,
+      "step": 5341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.680250130840386,
+      "learning_rate": 4.6414753674002696e-05,
+      "loss": 0.812,
+      "step": 5342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.285017795117825,
+      "learning_rate": 4.641260138254727e-05,
+      "loss": 0.9502,
+      "step": 5343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9323449003250575,
+      "learning_rate": 4.641044849518786e-05,
+      "loss": 0.8104,
+      "step": 5344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7364596022849934,
+      "learning_rate": 4.6408295011984385e-05,
+      "loss": 0.9619,
+      "step": 5345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8496277817046027,
+      "learning_rate": 4.6406140932996775e-05,
+      "loss": 1.2998,
+      "step": 5346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.7728470906992495,
+      "learning_rate": 4.640398625828498e-05,
+      "loss": 1.5137,
+      "step": 5347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.31688912782264,
+      "learning_rate": 4.6401830987908955e-05,
+      "loss": 1.0742,
+      "step": 5348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.202217614303418,
+      "learning_rate": 4.6399675121928686e-05,
+      "loss": 1.8906,
+      "step": 5349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.150380744473533,
+      "learning_rate": 4.639751866040418e-05,
+      "loss": 1.3086,
+      "step": 5350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8261110574543715,
+      "learning_rate": 4.639536160339543e-05,
+      "loss": 0.8848,
+      "step": 5351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4149072505481306,
+      "learning_rate": 4.6393203950962484e-05,
+      "loss": 0.8125,
+      "step": 5352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8631991694942616,
+      "learning_rate": 4.6391045703165384e-05,
+      "loss": 0.7998,
+      "step": 5353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4648021567533642,
+      "learning_rate": 4.638888686006418e-05,
+      "loss": 1.0063,
+      "step": 5354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.825857870893368,
+      "learning_rate": 4.638672742171897e-05,
+      "loss": 1.3564,
+      "step": 5355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.600957911722769,
+      "learning_rate": 4.638456738818984e-05,
+      "loss": 1.0566,
+      "step": 5356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9449303186481959,
+      "learning_rate": 4.63824067595369e-05,
+      "loss": 0.7515,
+      "step": 5357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8390272947008346,
+      "learning_rate": 4.6380245535820286e-05,
+      "loss": 0.9834,
+      "step": 5358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8932232700749054,
+      "learning_rate": 4.637808371710015e-05,
+      "loss": 0.871,
+      "step": 5359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6656317695366547,
+      "learning_rate": 4.637592130343664e-05,
+      "loss": 1.0723,
+      "step": 5360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8392036118085722,
+      "learning_rate": 4.6373758294889945e-05,
+      "loss": 1.1855,
+      "step": 5361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3362309198227997,
+      "learning_rate": 4.637159469152026e-05,
+      "loss": 0.915,
+      "step": 5362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6410997855514644,
+      "learning_rate": 4.6369430493387794e-05,
+      "loss": 1.4395,
+      "step": 5363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5586071683808447,
+      "learning_rate": 4.636726570055278e-05,
+      "loss": 1.4727,
+      "step": 5364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.107415276597724,
+      "learning_rate": 4.636510031307545e-05,
+      "loss": 0.8799,
+      "step": 5365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.862811492257789,
+      "learning_rate": 4.636293433101608e-05,
+      "loss": 1.1338,
+      "step": 5366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.287439996185,
+      "learning_rate": 4.6360767754434955e-05,
+      "loss": 1.2412,
+      "step": 5367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.319896313882482,
+      "learning_rate": 4.635860058339235e-05,
+      "loss": 1.0742,
+      "step": 5368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.121352037752955,
+      "learning_rate": 4.635643281794858e-05,
+      "loss": 0.8777,
+      "step": 5369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0100548242332446,
+      "learning_rate": 4.635426445816399e-05,
+      "loss": 0.8672,
+      "step": 5370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5997073228871077,
+      "learning_rate": 4.635209550409891e-05,
+      "loss": 1.4424,
+      "step": 5371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4769185769932855,
+      "learning_rate": 4.6349925955813706e-05,
+      "loss": 0.7783,
+      "step": 5372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6780658408255444,
+      "learning_rate": 4.6347755813368755e-05,
+      "loss": 0.7705,
+      "step": 5373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.80273497070706,
+      "learning_rate": 4.634558507682444e-05,
+      "loss": 0.96,
+      "step": 5374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6499482743595215,
+      "learning_rate": 4.634341374624119e-05,
+      "loss": 0.6831,
+      "step": 5375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.852026635144995,
+      "learning_rate": 4.634124182167942e-05,
+      "loss": 0.9185,
+      "step": 5376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0364594780674,
+      "learning_rate": 4.633906930319959e-05,
+      "loss": 0.54,
+      "step": 5377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5597192516460305,
+      "learning_rate": 4.633689619086215e-05,
+      "loss": 0.9854,
+      "step": 5378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.798920879605094,
+      "learning_rate": 4.6334722484727556e-05,
+      "loss": 1.1934,
+      "step": 5379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6942737291488754,
+      "learning_rate": 4.633254818485633e-05,
+      "loss": 1.4473,
+      "step": 5380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.289330032947838,
+      "learning_rate": 4.633037329130897e-05,
+      "loss": 1.834,
+      "step": 5381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9996137245996843,
+      "learning_rate": 4.632819780414601e-05,
+      "loss": 1.6426,
+      "step": 5382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.302791258842633,
+      "learning_rate": 4.6326021723427985e-05,
+      "loss": 0.8521,
+      "step": 5383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.236941164439253,
+      "learning_rate": 4.6323845049215455e-05,
+      "loss": 1.2637,
+      "step": 5384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.222816930772089,
+      "learning_rate": 4.632166778156901e-05,
+      "loss": 0.9951,
+      "step": 5385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.32864046358501,
+      "learning_rate": 4.631948992054922e-05,
+      "loss": 1.0039,
+      "step": 5386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.240040033187077,
+      "learning_rate": 4.631731146621671e-05,
+      "loss": 0.5908,
+      "step": 5387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.482915101473074,
+      "learning_rate": 4.631513241863209e-05,
+      "loss": 1.4473,
+      "step": 5388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.057078895350287,
+      "learning_rate": 4.631295277785602e-05,
+      "loss": 0.7866,
+      "step": 5389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3601849915043345,
+      "learning_rate": 4.631077254394916e-05,
+      "loss": 0.9756,
+      "step": 5390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3316456845011806,
+      "learning_rate": 4.630859171697216e-05,
+      "loss": 0.7515,
+      "step": 5391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.490935713583404,
+      "learning_rate": 4.6306410296985725e-05,
+      "loss": 0.7959,
+      "step": 5392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.636579894007622,
+      "learning_rate": 4.6304228284050574e-05,
+      "loss": 0.7603,
+      "step": 5393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.877210909250934,
+      "learning_rate": 4.630204567822742e-05,
+      "loss": 1.0146,
+      "step": 5394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.973957100943679,
+      "learning_rate": 4.6299862479577014e-05,
+      "loss": 1.3452,
+      "step": 5395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6973703683104238,
+      "learning_rate": 4.6297678688160096e-05,
+      "loss": 0.7134,
+      "step": 5396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8626316889826358,
+      "learning_rate": 4.6295494304037456e-05,
+      "loss": 1.3145,
+      "step": 5397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2170549822206604,
+      "learning_rate": 4.629330932726988e-05,
+      "loss": 0.7158,
+      "step": 5398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7060611192013515,
+      "learning_rate": 4.6291123757918174e-05,
+      "loss": 0.8672,
+      "step": 5399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.998590461415312,
+      "learning_rate": 4.628893759604316e-05,
+      "loss": 1.168,
+      "step": 5400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.951681895575553,
+      "learning_rate": 4.6286750841705674e-05,
+      "loss": 0.7979,
+      "step": 5401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5096515006487268,
+      "learning_rate": 4.628456349496659e-05,
+      "loss": 0.8428,
+      "step": 5402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.670650425869812,
+      "learning_rate": 4.628237555588676e-05,
+      "loss": 1.2344,
+      "step": 5403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0920855610565967,
+      "learning_rate": 4.628018702452708e-05,
+      "loss": 1.1152,
+      "step": 5404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.661822498786342,
+      "learning_rate": 4.627799790094847e-05,
+      "loss": 0.6426,
+      "step": 5405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.698620168116954,
+      "learning_rate": 4.627580818521183e-05,
+      "loss": 1.7324,
+      "step": 5406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.063186069013431,
+      "learning_rate": 4.627361787737811e-05,
+      "loss": 1.0195,
+      "step": 5407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.132154169931979,
+      "learning_rate": 4.627142697750827e-05,
+      "loss": 1.2676,
+      "step": 5408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.955919162940811,
+      "learning_rate": 4.626923548566328e-05,
+      "loss": 1.2974,
+      "step": 5409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.359175263891467,
+      "learning_rate": 4.626704340190411e-05,
+      "loss": 1.1982,
+      "step": 5410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0205123906265823,
+      "learning_rate": 4.626485072629179e-05,
+      "loss": 0.9805,
+      "step": 5411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.001724348329076,
+      "learning_rate": 4.626265745888733e-05,
+      "loss": 0.9678,
+      "step": 5412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.136296150731669,
+      "learning_rate": 4.626046359975177e-05,
+      "loss": 0.749,
+      "step": 5413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.856248709061714,
+      "learning_rate": 4.625826914894616e-05,
+      "loss": 0.8711,
+      "step": 5414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.721905947507277,
+      "learning_rate": 4.6256074106531586e-05,
+      "loss": 1.0283,
+      "step": 5415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8838942237640866,
+      "learning_rate": 4.6253878472569115e-05,
+      "loss": 1.2441,
+      "step": 5416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.120599318432466,
+      "learning_rate": 4.625168224711986e-05,
+      "loss": 1.3047,
+      "step": 5417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.193592308018352,
+      "learning_rate": 4.624948543024493e-05,
+      "loss": 0.9316,
+      "step": 5418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1973422295674228,
+      "learning_rate": 4.6247288022005476e-05,
+      "loss": 1.0068,
+      "step": 5419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8865409435578044,
+      "learning_rate": 4.6245090022462644e-05,
+      "loss": 0.7141,
+      "step": 5420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9511055233416033,
+      "learning_rate": 4.6242891431677615e-05,
+      "loss": 0.6141,
+      "step": 5421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5051977483852257,
+      "learning_rate": 4.624069224971156e-05,
+      "loss": 1.0151,
+      "step": 5422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.789468520907988,
+      "learning_rate": 4.623849247662568e-05,
+      "loss": 0.9434,
+      "step": 5423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.5035019909534615,
+      "learning_rate": 4.623629211248122e-05,
+      "loss": 0.9688,
+      "step": 5424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.751041483889162,
+      "learning_rate": 4.623409115733938e-05,
+      "loss": 1.2402,
+      "step": 5425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.10162456293604,
+      "learning_rate": 4.6231889611261425e-05,
+      "loss": 1.2109,
+      "step": 5426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.3167778580139435,
+      "learning_rate": 4.6229687474308635e-05,
+      "loss": 1.085,
+      "step": 5427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.922445063467044,
+      "learning_rate": 4.622748474654228e-05,
+      "loss": 1.8848,
+      "step": 5428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8191466302098849,
+      "learning_rate": 4.622528142802367e-05,
+      "loss": 1.2915,
+      "step": 5429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8142167215563465,
+      "learning_rate": 4.6223077518814116e-05,
+      "loss": 0.9509,
+      "step": 5430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6035308265099153,
+      "learning_rate": 4.622087301897495e-05,
+      "loss": 1.2227,
+      "step": 5431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0942563256064264,
+      "learning_rate": 4.621866792856754e-05,
+      "loss": 1.2227,
+      "step": 5432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.061113180289154,
+      "learning_rate": 4.6216462247653226e-05,
+      "loss": 0.8115,
+      "step": 5433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1855920372763444,
+      "learning_rate": 4.621425597629341e-05,
+      "loss": 1.0591,
+      "step": 5434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.142174542384132,
+      "learning_rate": 4.621204911454949e-05,
+      "loss": 1.3379,
+      "step": 5435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5861650116331263,
+      "learning_rate": 4.620984166248288e-05,
+      "loss": 1.0596,
+      "step": 5436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.892240312377127,
+      "learning_rate": 4.6207633620155e-05,
+      "loss": 1.5234,
+      "step": 5437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5861485328885068,
+      "learning_rate": 4.620542498762732e-05,
+      "loss": 1.3818,
+      "step": 5438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.595500088135461,
+      "learning_rate": 4.6203215764961295e-05,
+      "loss": 2.0195,
+      "step": 5439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0676820562760483,
+      "learning_rate": 4.62010059522184e-05,
+      "loss": 1.0068,
+      "step": 5440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.56285682363571,
+      "learning_rate": 4.6198795549460144e-05,
+      "loss": 1.1152,
+      "step": 5441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3531009726538956,
+      "learning_rate": 4.6196584556748035e-05,
+      "loss": 0.5874,
+      "step": 5442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.725665036144276,
+      "learning_rate": 4.6194372974143613e-05,
+      "loss": 0.9729,
+      "step": 5443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0478431141744493,
+      "learning_rate": 4.6192160801708416e-05,
+      "loss": 1.0859,
+      "step": 5444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.7602434551923185,
+      "learning_rate": 4.618994803950402e-05,
+      "loss": 1.3047,
+      "step": 5445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0020085604530995,
+      "learning_rate": 4.6187734687591976e-05,
+      "loss": 0.5933,
+      "step": 5446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.296586102852171,
+      "learning_rate": 4.618552074603392e-05,
+      "loss": 0.8779,
+      "step": 5447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.160545200964092,
+      "learning_rate": 4.6183306214891444e-05,
+      "loss": 1.2129,
+      "step": 5448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.1789954249378995,
+      "learning_rate": 4.618109109422617e-05,
+      "loss": 1.021,
+      "step": 5449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.374825585662659,
+      "learning_rate": 4.617887538409976e-05,
+      "loss": 2.3945,
+      "step": 5450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7519674821518394,
+      "learning_rate": 4.617665908457387e-05,
+      "loss": 1.2617,
+      "step": 5451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.916391375749686,
+      "learning_rate": 4.617444219571018e-05,
+      "loss": 1.3418,
+      "step": 5452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.719015525041194,
+      "learning_rate": 4.6172224717570376e-05,
+      "loss": 0.9961,
+      "step": 5453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6802124754815253,
+      "learning_rate": 4.617000665021618e-05,
+      "loss": 0.9902,
+      "step": 5454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.913970225987883,
+      "learning_rate": 4.616778799370932e-05,
+      "loss": 1.5977,
+      "step": 5455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.626781435846094,
+      "learning_rate": 4.6165568748111544e-05,
+      "loss": 0.9404,
+      "step": 5456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.786954797231037,
+      "learning_rate": 4.6163348913484597e-05,
+      "loss": 1.2126,
+      "step": 5457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.467986362994586,
+      "learning_rate": 4.616112848989027e-05,
+      "loss": 1.377,
+      "step": 5458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.259866169874537,
+      "learning_rate": 4.615890747739035e-05,
+      "loss": 1.2246,
+      "step": 5459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1124634914516425,
+      "learning_rate": 4.615668587604665e-05,
+      "loss": 1.053,
+      "step": 5460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2207826786335927,
+      "learning_rate": 4.615446368592099e-05,
+      "loss": 1.3887,
+      "step": 5461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.060913822278287,
+      "learning_rate": 4.615224090707522e-05,
+      "loss": 1.6309,
+      "step": 5462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.484446667886045,
+      "learning_rate": 4.61500175395712e-05,
+      "loss": 0.9126,
+      "step": 5463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8779658071622087,
+      "learning_rate": 4.61477935834708e-05,
+      "loss": 1.2402,
+      "step": 5464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.214402573661742,
+      "learning_rate": 4.6145569038835914e-05,
+      "loss": 0.8945,
+      "step": 5465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0725376789165715,
+      "learning_rate": 4.6143343905728455e-05,
+      "loss": 1.0723,
+      "step": 5466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.298265932103491,
+      "learning_rate": 4.614111818421033e-05,
+      "loss": 1.7129,
+      "step": 5467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.506399055375485,
+      "learning_rate": 4.613889187434351e-05,
+      "loss": 1.4707,
+      "step": 5468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.261957250896923,
+      "learning_rate": 4.6136664976189917e-05,
+      "loss": 0.8828,
+      "step": 5469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.05041160701256,
+      "learning_rate": 4.613443748981155e-05,
+      "loss": 1.5215,
+      "step": 5470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.959278970831946,
+      "learning_rate": 4.6132209415270394e-05,
+      "loss": 0.4165,
+      "step": 5471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.004483772849917,
+      "learning_rate": 4.612998075262845e-05,
+      "loss": 1.3389,
+      "step": 5472
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0805536748454028,
+      "learning_rate": 4.612775150194775e-05,
+      "loss": 0.9004,
+      "step": 5473
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.503826774112596,
+      "learning_rate": 4.612552166329031e-05,
+      "loss": 0.8933,
+      "step": 5474
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.294672731092266,
+      "learning_rate": 4.612329123671822e-05,
+      "loss": 1.416,
+      "step": 5475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6119426954213676,
+      "learning_rate": 4.612106022229352e-05,
+      "loss": 1.125,
+      "step": 5476
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.022158921292006,
+      "learning_rate": 4.611882862007832e-05,
+      "loss": 1.4014,
+      "step": 5477
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6913066784692603,
+      "learning_rate": 4.611659643013472e-05,
+      "loss": 0.8291,
+      "step": 5478
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.322139636451754,
+      "learning_rate": 4.611436365252484e-05,
+      "loss": 1.3066,
+      "step": 5479
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5843145803932908,
+      "learning_rate": 4.611213028731081e-05,
+      "loss": 1.2988,
+      "step": 5480
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.89322314077781,
+      "learning_rate": 4.6109896334554785e-05,
+      "loss": 1.4023,
+      "step": 5481
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.676517693087951,
+      "learning_rate": 4.610766179431895e-05,
+      "loss": 1.1729,
+      "step": 5482
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3946350220264225,
+      "learning_rate": 4.6105426666665465e-05,
+      "loss": 0.9326,
+      "step": 5483
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5343231062448064,
+      "learning_rate": 4.610319095165656e-05,
+      "loss": 1.375,
+      "step": 5484
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.692378239624463,
+      "learning_rate": 4.6100954649354436e-05,
+      "loss": 0.9009,
+      "step": 5485
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4409960914773645,
+      "learning_rate": 4.609871775982134e-05,
+      "loss": 0.5479,
+      "step": 5486
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.140280937213106,
+      "learning_rate": 4.6096480283119516e-05,
+      "loss": 1.4238,
+      "step": 5487
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.504824516416756,
+      "learning_rate": 4.609424221931122e-05,
+      "loss": 0.6838,
+      "step": 5488
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.415616421110605,
+      "learning_rate": 4.6092003568458766e-05,
+      "loss": 1.1064,
+      "step": 5489
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.029483851848111,
+      "learning_rate": 4.608976433062444e-05,
+      "loss": 1.4141,
+      "step": 5490
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2547034799031995,
+      "learning_rate": 4.608752450587055e-05,
+      "loss": 1.0879,
+      "step": 5491
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.718034096278086,
+      "learning_rate": 4.6085284094259446e-05,
+      "loss": 1.0889,
+      "step": 5492
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.141784233980616,
+      "learning_rate": 4.6083043095853464e-05,
+      "loss": 1.5254,
+      "step": 5493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.274591767746224,
+      "learning_rate": 4.6080801510714975e-05,
+      "loss": 1.1719,
+      "step": 5494
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.408097492318974,
+      "learning_rate": 4.607855933890637e-05,
+      "loss": 0.9072,
+      "step": 5495
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7816288193886778,
+      "learning_rate": 4.607631658049003e-05,
+      "loss": 0.6777,
+      "step": 5496
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.070938236318563,
+      "learning_rate": 4.607407323552838e-05,
+      "loss": 1.7812,
+      "step": 5497
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.282792530479484,
+      "learning_rate": 4.607182930408386e-05,
+      "loss": 1.1113,
+      "step": 5498
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.130174598214646,
+      "learning_rate": 4.60695847862189e-05,
+      "loss": 1.9688,
+      "step": 5499
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.035513760435163,
+      "learning_rate": 4.606733968199598e-05,
+      "loss": 0.8066,
+      "step": 5500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7281287648807515,
+      "learning_rate": 4.606509399147757e-05,
+      "loss": 1.2881,
+      "step": 5501
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.209264834712533,
+      "learning_rate": 4.606284771472617e-05,
+      "loss": 0.5676,
+      "step": 5502
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8501011356855643,
+      "learning_rate": 4.6060600851804295e-05,
+      "loss": 0.6857,
+      "step": 5503
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6290538910761447,
+      "learning_rate": 4.6058353402774464e-05,
+      "loss": 1.002,
+      "step": 5504
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0468347791315638,
+      "learning_rate": 4.605610536769923e-05,
+      "loss": 1.1162,
+      "step": 5505
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.512381031768769,
+      "learning_rate": 4.605385674664117e-05,
+      "loss": 0.311,
+      "step": 5506
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.790399626143225,
+      "learning_rate": 4.6051607539662835e-05,
+      "loss": 0.877,
+      "step": 5507
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3951005118986486,
+      "learning_rate": 4.604935774682684e-05,
+      "loss": 0.9336,
+      "step": 5508
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.054516919383652,
+      "learning_rate": 4.6047107368195784e-05,
+      "loss": 0.7825,
+      "step": 5509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8980972049780913,
+      "learning_rate": 4.60448564038323e-05,
+      "loss": 0.5801,
+      "step": 5510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2162876525169373,
+      "learning_rate": 4.604260485379903e-05,
+      "loss": 1.0381,
+      "step": 5511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2097733143657137,
+      "learning_rate": 4.604035271815863e-05,
+      "loss": 0.5396,
+      "step": 5512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9912619491446883,
+      "learning_rate": 4.603809999697378e-05,
+      "loss": 1.4434,
+      "step": 5513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.711327076823652,
+      "learning_rate": 4.603584669030717e-05,
+      "loss": 1.3809,
+      "step": 5514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3626296477090705,
+      "learning_rate": 4.603359279822151e-05,
+      "loss": 1.3008,
+      "step": 5515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1306580248534175,
+      "learning_rate": 4.6031338320779534e-05,
+      "loss": 1.502,
+      "step": 5516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.190450803439794,
+      "learning_rate": 4.602908325804397e-05,
+      "loss": 0.8027,
+      "step": 5517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.431866790931726,
+      "learning_rate": 4.602682761007757e-05,
+      "loss": 1.25,
+      "step": 5518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.044644361801984,
+      "learning_rate": 4.602457137694313e-05,
+      "loss": 1.1704,
+      "step": 5519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.267198322280141,
+      "learning_rate": 4.6022314558703426e-05,
+      "loss": 1.3867,
+      "step": 5520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2773174152975773,
+      "learning_rate": 4.602005715542126e-05,
+      "loss": 1.2188,
+      "step": 5521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.416987945861177,
+      "learning_rate": 4.601779916715947e-05,
+      "loss": 1.1094,
+      "step": 5522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.130958826837875,
+      "learning_rate": 4.6015540593980875e-05,
+      "loss": 1.0332,
+      "step": 5523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1884921156779926,
+      "learning_rate": 4.601328143594835e-05,
+      "loss": 0.479,
+      "step": 5524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.40382778617144,
+      "learning_rate": 4.601102169312475e-05,
+      "loss": 0.7803,
+      "step": 5525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.716841788736449,
+      "learning_rate": 4.6008761365572975e-05,
+      "loss": 1.5317,
+      "step": 5526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.245904586124073,
+      "learning_rate": 4.600650045335592e-05,
+      "loss": 0.5337,
+      "step": 5527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.22305972261218,
+      "learning_rate": 4.600423895653651e-05,
+      "loss": 1.1143,
+      "step": 5528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9397086099000704,
+      "learning_rate": 4.600197687517769e-05,
+      "loss": 0.8105,
+      "step": 5529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.958046148177152,
+      "learning_rate": 4.599971420934239e-05,
+      "loss": 1.1553,
+      "step": 5530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.675707823671593,
+      "learning_rate": 4.5997450959093595e-05,
+      "loss": 0.8789,
+      "step": 5531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.540172342734733,
+      "learning_rate": 4.599518712449429e-05,
+      "loss": 1.2168,
+      "step": 5532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.231765449716127,
+      "learning_rate": 4.599292270560748e-05,
+      "loss": 1.0996,
+      "step": 5533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6389544919431396,
+      "learning_rate": 4.599065770249616e-05,
+      "loss": 1.0137,
+      "step": 5534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.178460207341421,
+      "learning_rate": 4.59883921152234e-05,
+      "loss": 1.3105,
+      "step": 5535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.699927406748001,
+      "learning_rate": 4.598612594385223e-05,
+      "loss": 0.5962,
+      "step": 5536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.332368385919095,
+      "learning_rate": 4.598385918844571e-05,
+      "loss": 1.25,
+      "step": 5537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3809815136263905,
+      "learning_rate": 4.598159184906693e-05,
+      "loss": 0.6885,
+      "step": 5538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.652577438630211,
+      "learning_rate": 4.5979323925778997e-05,
+      "loss": 0.7822,
+      "step": 5539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7851474328456596,
+      "learning_rate": 4.597705541864502e-05,
+      "loss": 1.2285,
+      "step": 5540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3026653146782143,
+      "learning_rate": 4.5974786327728134e-05,
+      "loss": 1.3164,
+      "step": 5541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.865427212843081,
+      "learning_rate": 4.597251665309148e-05,
+      "loss": 1.2676,
+      "step": 5542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.093664936216591,
+      "learning_rate": 4.597024639479822e-05,
+      "loss": 1.0898,
+      "step": 5543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.297804631326074,
+      "learning_rate": 4.596797555291155e-05,
+      "loss": 0.9327,
+      "step": 5544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.701805813329766,
+      "learning_rate": 4.5965704127494644e-05,
+      "loss": 1.2441,
+      "step": 5545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0569713359239463,
+      "learning_rate": 4.596343211861073e-05,
+      "loss": 1.0327,
+      "step": 5546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.129108810147783,
+      "learning_rate": 4.596115952632304e-05,
+      "loss": 0.7471,
+      "step": 5547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.817109513741909,
+      "learning_rate": 4.595888635069481e-05,
+      "loss": 1.0781,
+      "step": 5548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.650008131044293,
+      "learning_rate": 4.5956612591789305e-05,
+      "loss": 1.5352,
+      "step": 5549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.430527608729545,
+      "learning_rate": 4.5954338249669804e-05,
+      "loss": 1.8496,
+      "step": 5550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.906261937468162,
+      "learning_rate": 4.59520633243996e-05,
+      "loss": 2.2402,
+      "step": 5551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.552150914375525,
+      "learning_rate": 4.594978781604201e-05,
+      "loss": 1.4561,
+      "step": 5552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.363275854624784,
+      "learning_rate": 4.594751172466034e-05,
+      "loss": 1.2734,
+      "step": 5553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.410814160142484,
+      "learning_rate": 4.5945235050317956e-05,
+      "loss": 1.9941,
+      "step": 5554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.899423601154737,
+      "learning_rate": 4.594295779307821e-05,
+      "loss": 1.207,
+      "step": 5555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9045951808070734,
+      "learning_rate": 4.594067995300447e-05,
+      "loss": 1.4258,
+      "step": 5556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5718515363641794,
+      "learning_rate": 4.593840153016013e-05,
+      "loss": 0.9043,
+      "step": 5557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.300821456169284,
+      "learning_rate": 4.59361225246086e-05,
+      "loss": 1.4629,
+      "step": 5558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.706117946583318,
+      "learning_rate": 4.5933842936413306e-05,
+      "loss": 1.1992,
+      "step": 5559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4457705792236886,
+      "learning_rate": 4.5931562765637684e-05,
+      "loss": 1.0635,
+      "step": 5560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.61826197740483,
+      "learning_rate": 4.5929282012345185e-05,
+      "loss": 1.7148,
+      "step": 5561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.410613135040271,
+      "learning_rate": 4.5927000676599296e-05,
+      "loss": 1.6387,
+      "step": 5562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.905313976870524,
+      "learning_rate": 4.5924718758463496e-05,
+      "loss": 1.2109,
+      "step": 5563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.248117200112643,
+      "learning_rate": 4.592243625800129e-05,
+      "loss": 1.0049,
+      "step": 5564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.6269568536303165,
+      "learning_rate": 4.59201531752762e-05,
+      "loss": 1.0566,
+      "step": 5565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2906787361201566,
+      "learning_rate": 4.591786951035177e-05,
+      "loss": 0.3491,
+      "step": 5566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.402601300745719,
+      "learning_rate": 4.5915585263291535e-05,
+      "loss": 1.4707,
+      "step": 5567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.635715412401054,
+      "learning_rate": 4.591330043415909e-05,
+      "loss": 0.9863,
+      "step": 5568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2333596863781287,
+      "learning_rate": 4.5911015023018e-05,
+      "loss": 0.7891,
+      "step": 5569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.679476715783973,
+      "learning_rate": 4.590872902993188e-05,
+      "loss": 0.6414,
+      "step": 5570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8298784249633826,
+      "learning_rate": 4.590644245496434e-05,
+      "loss": 1.3594,
+      "step": 5571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.842117163047772,
+      "learning_rate": 4.590415529817902e-05,
+      "loss": 1.1982,
+      "step": 5572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3811606763371564,
+      "learning_rate": 4.590186755963956e-05,
+      "loss": 1.2261,
+      "step": 5573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.77432764899463,
+      "learning_rate": 4.589957923940964e-05,
+      "loss": 1.4141,
+      "step": 5574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.79230581756058,
+      "learning_rate": 4.589729033755294e-05,
+      "loss": 0.98,
+      "step": 5575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1074223275312303,
+      "learning_rate": 4.589500085413315e-05,
+      "loss": 0.9888,
+      "step": 5576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.900474734633924,
+      "learning_rate": 4.5892710789214e-05,
+      "loss": 0.8081,
+      "step": 5577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.190685825347395,
+      "learning_rate": 4.589042014285921e-05,
+      "loss": 1.04,
+      "step": 5578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.132275188800434,
+      "learning_rate": 4.588812891513254e-05,
+      "loss": 0.9355,
+      "step": 5579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9400828588625587,
+      "learning_rate": 4.588583710609774e-05,
+      "loss": 1.123,
+      "step": 5580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.608863166749779,
+      "learning_rate": 4.5883544715818595e-05,
+      "loss": 1.0781,
+      "step": 5581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.556370628654353,
+      "learning_rate": 4.58812517443589e-05,
+      "loss": 1.3203,
+      "step": 5582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.216812253908964,
+      "learning_rate": 4.587895819178248e-05,
+      "loss": 1.3057,
+      "step": 5583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.59131848448567,
+      "learning_rate": 4.587666405815314e-05,
+      "loss": 0.9941,
+      "step": 5584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.260019097239126,
+      "learning_rate": 4.587436934353475e-05,
+      "loss": 1.2324,
+      "step": 5585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6287684091067525,
+      "learning_rate": 4.587207404799115e-05,
+      "loss": 0.8994,
+      "step": 5586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.66790562032364,
+      "learning_rate": 4.586977817158623e-05,
+      "loss": 1.1758,
+      "step": 5587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4300883167999285,
+      "learning_rate": 4.586748171438388e-05,
+      "loss": 1.0762,
+      "step": 5588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2326990475482007,
+      "learning_rate": 4.586518467644801e-05,
+      "loss": 1.1484,
+      "step": 5589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.920666936935282,
+      "learning_rate": 4.586288705784254e-05,
+      "loss": 1.5645,
+      "step": 5590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.316335769008847,
+      "learning_rate": 4.5860588858631423e-05,
+      "loss": 1.0723,
+      "step": 5591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.022478837549697,
+      "learning_rate": 4.585829007887861e-05,
+      "loss": 0.8916,
+      "step": 5592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.024069845636205,
+      "learning_rate": 4.585599071864807e-05,
+      "loss": 1.4883,
+      "step": 5593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.918065326118927,
+      "learning_rate": 4.58536907780038e-05,
+      "loss": 0.7842,
+      "step": 5594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3138358823280436,
+      "learning_rate": 4.5851390257009805e-05,
+      "loss": 1.1562,
+      "step": 5595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.261050447362426,
+      "learning_rate": 4.584908915573011e-05,
+      "loss": 1.2217,
+      "step": 5596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7272694717734587,
+      "learning_rate": 4.584678747422875e-05,
+      "loss": 1.7773,
+      "step": 5597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6985247718547551,
+      "learning_rate": 4.5844485212569786e-05,
+      "loss": 1.0576,
+      "step": 5598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0767169283585583,
+      "learning_rate": 4.584218237081728e-05,
+      "loss": 1.5137,
+      "step": 5599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9601293471532462,
+      "learning_rate": 4.583987894903532e-05,
+      "loss": 0.8438,
+      "step": 5600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.959247416643872,
+      "learning_rate": 4.583757494728802e-05,
+      "loss": 1.1641,
+      "step": 5601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8603037175389927,
+      "learning_rate": 4.583527036563949e-05,
+      "loss": 0.8057,
+      "step": 5602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6111518482616614,
+      "learning_rate": 4.5832965204153864e-05,
+      "loss": 0.6494,
+      "step": 5603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.586143995896622,
+      "learning_rate": 4.5830659462895296e-05,
+      "loss": 1.1035,
+      "step": 5604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.999641977843635,
+      "learning_rate": 4.582835314192797e-05,
+      "loss": 1.3691,
+      "step": 5605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.003857468872025,
+      "learning_rate": 4.582604624131603e-05,
+      "loss": 1.2676,
+      "step": 5606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.438212559547397,
+      "learning_rate": 4.5823738761123725e-05,
+      "loss": 0.6377,
+      "step": 5607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.146837305867917,
+      "learning_rate": 4.5821430701415234e-05,
+      "loss": 1.0195,
+      "step": 5608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4901435212625898,
+      "learning_rate": 4.581912206225482e-05,
+      "loss": 1.0889,
+      "step": 5609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3013494596273616,
+      "learning_rate": 4.5816812843706694e-05,
+      "loss": 1.2461,
+      "step": 5610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.0217865267297,
+      "learning_rate": 4.581450304583515e-05,
+      "loss": 1.3535,
+      "step": 5611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.687899892464223,
+      "learning_rate": 4.581219266870446e-05,
+      "loss": 1.2441,
+      "step": 5612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.571180852053878,
+      "learning_rate": 4.580988171237892e-05,
+      "loss": 0.9175,
+      "step": 5613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9107845575407123,
+      "learning_rate": 4.5807570176922844e-05,
+      "loss": 1.5547,
+      "step": 5614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4852421113558334,
+      "learning_rate": 4.5805258062400557e-05,
+      "loss": 1.3682,
+      "step": 5615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.656793754625613,
+      "learning_rate": 4.5802945368876416e-05,
+      "loss": 1.3887,
+      "step": 5616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.888333788954458,
+      "learning_rate": 4.580063209641476e-05,
+      "loss": 1.1523,
+      "step": 5617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7721101405238335,
+      "learning_rate": 4.579831824508e-05,
+      "loss": 0.8076,
+      "step": 5618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.3520231373170954,
+      "learning_rate": 4.579600381493649e-05,
+      "loss": 1.0547,
+      "step": 5619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.694328725518435,
+      "learning_rate": 4.579368880604867e-05,
+      "loss": 1.4492,
+      "step": 5620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9524100916753198,
+      "learning_rate": 4.579137321848096e-05,
+      "loss": 1.1797,
+      "step": 5621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2797207129597314,
+      "learning_rate": 4.578905705229779e-05,
+      "loss": 1.0547,
+      "step": 5622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4479402797993983,
+      "learning_rate": 4.5786740307563636e-05,
+      "loss": 1.5566,
+      "step": 5623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.813876430073737,
+      "learning_rate": 4.578442298434295e-05,
+      "loss": 1.6729,
+      "step": 5624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6744839607252757,
+      "learning_rate": 4.578210508270024e-05,
+      "loss": 1.3613,
+      "step": 5625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9737493883655843,
+      "learning_rate": 4.5779786602700005e-05,
+      "loss": 1.002,
+      "step": 5626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0817538331163385,
+      "learning_rate": 4.577746754440677e-05,
+      "loss": 1.4082,
+      "step": 5627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.405619154635845,
+      "learning_rate": 4.5775147907885065e-05,
+      "loss": 0.7178,
+      "step": 5628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.087086153224195,
+      "learning_rate": 4.5772827693199455e-05,
+      "loss": 1.8301,
+      "step": 5629
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.084518874494682,
+      "learning_rate": 4.57705069004145e-05,
+      "loss": 1.2148,
+      "step": 5630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4620388508891575,
+      "learning_rate": 4.5768185529594806e-05,
+      "loss": 1.1113,
+      "step": 5631
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.988390966359815,
+      "learning_rate": 4.5765863580804956e-05,
+      "loss": 1.1489,
+      "step": 5632
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.323288367487122,
+      "learning_rate": 4.5763541054109585e-05,
+      "loss": 1.1328,
+      "step": 5633
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.922574046349636,
+      "learning_rate": 4.576121794957331e-05,
+      "loss": 1.1816,
+      "step": 5634
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4912630955743813,
+      "learning_rate": 4.57588942672608e-05,
+      "loss": 1.4121,
+      "step": 5635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.787447079553884,
+      "learning_rate": 4.57565700072367e-05,
+      "loss": 0.8154,
+      "step": 5636
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5634309891782334,
+      "learning_rate": 4.5754245169565716e-05,
+      "loss": 0.7725,
+      "step": 5637
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.46145666076623,
+      "learning_rate": 4.575191975431254e-05,
+      "loss": 2.3203,
+      "step": 5638
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1999165166455477,
+      "learning_rate": 4.574959376154188e-05,
+      "loss": 1.252,
+      "step": 5639
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.313707959150534,
+      "learning_rate": 4.5747267191318476e-05,
+      "loss": 0.6841,
+      "step": 5640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1315926776937033,
+      "learning_rate": 4.574494004370707e-05,
+      "loss": 0.9082,
+      "step": 5641
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7623047807163865,
+      "learning_rate": 4.5742612318772424e-05,
+      "loss": 0.6821,
+      "step": 5642
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.0706615920657345,
+      "learning_rate": 4.5740284016579324e-05,
+      "loss": 1.1914,
+      "step": 5643
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0541654565105323,
+      "learning_rate": 4.573795513719257e-05,
+      "loss": 1.6133,
+      "step": 5644
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.0798046409360245,
+      "learning_rate": 4.573562568067696e-05,
+      "loss": 0.9629,
+      "step": 5645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8343875651631683,
+      "learning_rate": 4.5733295647097336e-05,
+      "loss": 0.8994,
+      "step": 5646
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2143882538745787,
+      "learning_rate": 4.573096503651852e-05,
+      "loss": 1.0669,
+      "step": 5647
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.409324792707098,
+      "learning_rate": 4.57286338490054e-05,
+      "loss": 1.0762,
+      "step": 5648
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.223858192289951,
+      "learning_rate": 4.572630208462284e-05,
+      "loss": 1.5449,
+      "step": 5649
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7187495615290147,
+      "learning_rate": 4.572396974343572e-05,
+      "loss": 1.0547,
+      "step": 5650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.900939605462753,
+      "learning_rate": 4.572163682550897e-05,
+      "loss": 1.1289,
+      "step": 5651
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7913245301761274,
+      "learning_rate": 4.57193033309075e-05,
+      "loss": 1.3164,
+      "step": 5652
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.273914185631233,
+      "learning_rate": 4.5716969259696254e-05,
+      "loss": 0.5654,
+      "step": 5653
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.733208943913028,
+      "learning_rate": 4.5714634611940177e-05,
+      "loss": 1.2656,
+      "step": 5654
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.003491308529432,
+      "learning_rate": 4.571229938770426e-05,
+      "loss": 0.6396,
+      "step": 5655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7576014723182216,
+      "learning_rate": 4.5709963587053484e-05,
+      "loss": 1.5781,
+      "step": 5656
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.957846936550231,
+      "learning_rate": 4.570762721005285e-05,
+      "loss": 2.0488,
+      "step": 5657
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.088034388207402,
+      "learning_rate": 4.5705290256767376e-05,
+      "loss": 1.1128,
+      "step": 5658
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.5502747085979385,
+      "learning_rate": 4.570295272726211e-05,
+      "loss": 1.6431,
+      "step": 5659
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6756938391003118,
+      "learning_rate": 4.5700614621602093e-05,
+      "loss": 0.7471,
+      "step": 5660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.025303709047251,
+      "learning_rate": 4.569827593985241e-05,
+      "loss": 1.3164,
+      "step": 5661
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7152780748016463,
+      "learning_rate": 4.569593668207812e-05,
+      "loss": 1.3438,
+      "step": 5662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.121217053168207,
+      "learning_rate": 4.5693596848344345e-05,
+      "loss": 1.2988,
+      "step": 5663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8947695238169786,
+      "learning_rate": 4.569125643871619e-05,
+      "loss": 0.9053,
+      "step": 5664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.109832375001447,
+      "learning_rate": 4.568891545325879e-05,
+      "loss": 1.4355,
+      "step": 5665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.09351054968249,
+      "learning_rate": 4.56865738920373e-05,
+      "loss": 2.167,
+      "step": 5666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.43632147814451,
+      "learning_rate": 4.568423175511688e-05,
+      "loss": 0.9199,
+      "step": 5667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.317158397966248,
+      "learning_rate": 4.568188904256271e-05,
+      "loss": 1.502,
+      "step": 5668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.321210116092896,
+      "learning_rate": 4.567954575443999e-05,
+      "loss": 0.8418,
+      "step": 5669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8397285616468713,
+      "learning_rate": 4.5677201890813925e-05,
+      "loss": 1.127,
+      "step": 5670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8609184727017807,
+      "learning_rate": 4.567485745174975e-05,
+      "loss": 0.8916,
+      "step": 5671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.094213739953668,
+      "learning_rate": 4.567251243731271e-05,
+      "loss": 1.498,
+      "step": 5672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.589890292317215,
+      "learning_rate": 4.5670166847568064e-05,
+      "loss": 1.168,
+      "step": 5673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.189204765729872,
+      "learning_rate": 4.566782068258109e-05,
+      "loss": 1.0859,
+      "step": 5674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1387795867193904,
+      "learning_rate": 4.566547394241708e-05,
+      "loss": 1.0664,
+      "step": 5675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.145283749975332,
+      "learning_rate": 4.566312662714135e-05,
+      "loss": 0.5181,
+      "step": 5676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.702898188534604,
+      "learning_rate": 4.566077873681921e-05,
+      "loss": 1.5098,
+      "step": 5677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6466327916954104,
+      "learning_rate": 4.5658430271516e-05,
+      "loss": 0.6938,
+      "step": 5678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1713731490880783,
+      "learning_rate": 4.5656081231297104e-05,
+      "loss": 1.0166,
+      "step": 5679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4338367710321895,
+      "learning_rate": 4.565373161622787e-05,
+      "loss": 1.0371,
+      "step": 5680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 22.594579320014734,
+      "learning_rate": 4.565138142637369e-05,
+      "loss": 0.8571,
+      "step": 5681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.944692851764243,
+      "learning_rate": 4.564903066179997e-05,
+      "loss": 1.4961,
+      "step": 5682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1847894496426927,
+      "learning_rate": 4.5646679322572136e-05,
+      "loss": 0.916,
+      "step": 5683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6249634313383132,
+      "learning_rate": 4.564432740875563e-05,
+      "loss": 0.7876,
+      "step": 5684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.051076929062806,
+      "learning_rate": 4.564197492041589e-05,
+      "loss": 1.0598,
+      "step": 5685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.044761076143551,
+      "learning_rate": 4.5639621857618386e-05,
+      "loss": 1.1445,
+      "step": 5686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.063244787464705,
+      "learning_rate": 4.563726822042861e-05,
+      "loss": 0.9199,
+      "step": 5687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.96845586693029,
+      "learning_rate": 4.5634914008912065e-05,
+      "loss": 0.9639,
+      "step": 5688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8796661339492973,
+      "learning_rate": 4.5632559223134264e-05,
+      "loss": 1.4336,
+      "step": 5689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.668599580550345,
+      "learning_rate": 4.5630203863160736e-05,
+      "loss": 0.8745,
+      "step": 5690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.722206280304431,
+      "learning_rate": 4.5627847929057035e-05,
+      "loss": 1.0254,
+      "step": 5691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7128691775501665,
+      "learning_rate": 4.562549142088872e-05,
+      "loss": 1.2988,
+      "step": 5692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.670480449572387,
+      "learning_rate": 4.562313433872139e-05,
+      "loss": 0.5691,
+      "step": 5693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.9656037740091685,
+      "learning_rate": 4.562077668262061e-05,
+      "loss": 1.0244,
+      "step": 5694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4785226471593296,
+      "learning_rate": 4.5618418452652014e-05,
+      "loss": 0.957,
+      "step": 5695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.141621259309806,
+      "learning_rate": 4.5616059648881224e-05,
+      "loss": 1.0107,
+      "step": 5696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.699309058839732,
+      "learning_rate": 4.561370027137389e-05,
+      "loss": 1.2637,
+      "step": 5697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.970909944739589,
+      "learning_rate": 4.561134032019567e-05,
+      "loss": 2.002,
+      "step": 5698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.2543052096432294,
+      "learning_rate": 4.5608979795412245e-05,
+      "loss": 1.3643,
+      "step": 5699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3638563228843834,
+      "learning_rate": 4.560661869708929e-05,
+      "loss": 1.1309,
+      "step": 5700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.334682963568769,
+      "learning_rate": 4.560425702529254e-05,
+      "loss": 0.9673,
+      "step": 5701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.040683297495617,
+      "learning_rate": 4.5601894780087696e-05,
+      "loss": 1.0518,
+      "step": 5702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2762294549240027,
+      "learning_rate": 4.5599531961540507e-05,
+      "loss": 1.207,
+      "step": 5703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.116005797477581,
+      "learning_rate": 4.559716856971673e-05,
+      "loss": 0.833,
+      "step": 5704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3991092412620696,
+      "learning_rate": 4.559480460468214e-05,
+      "loss": 1.1621,
+      "step": 5705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.121909755596306,
+      "learning_rate": 4.559244006650252e-05,
+      "loss": 0.7666,
+      "step": 5706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.884041211620796,
+      "learning_rate": 4.5590074955243674e-05,
+      "loss": 0.7529,
+      "step": 5707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.022474134273203,
+      "learning_rate": 4.5587709270971425e-05,
+      "loss": 0.7842,
+      "step": 5708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.016527601396301,
+      "learning_rate": 4.558534301375162e-05,
+      "loss": 0.8594,
+      "step": 5709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1548664242797133,
+      "learning_rate": 4.558297618365008e-05,
+      "loss": 0.8418,
+      "step": 5710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9809594985310923,
+      "learning_rate": 4.55806087807327e-05,
+      "loss": 0.8745,
+      "step": 5711
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.258615383600501,
+      "learning_rate": 4.557824080506535e-05,
+      "loss": 1.8184,
+      "step": 5712
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.097967872370817,
+      "learning_rate": 4.557587225671395e-05,
+      "loss": 1.3301,
+      "step": 5713
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.30716253392276,
+      "learning_rate": 4.557350313574439e-05,
+      "loss": 1.2109,
+      "step": 5714
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.252265873896868,
+      "learning_rate": 4.557113344222262e-05,
+      "loss": 0.9766,
+      "step": 5715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9430784459704338,
+      "learning_rate": 4.556876317621458e-05,
+      "loss": 1.4453,
+      "step": 5716
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.822538897209997,
+      "learning_rate": 4.556639233778623e-05,
+      "loss": 1.0859,
+      "step": 5717
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.545504670800266,
+      "learning_rate": 4.556402092700356e-05,
+      "loss": 0.752,
+      "step": 5718
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.266611450121152,
+      "learning_rate": 4.556164894393255e-05,
+      "loss": 1.3359,
+      "step": 5719
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6015144394586835,
+      "learning_rate": 4.5559276388639225e-05,
+      "loss": 1.502,
+      "step": 5720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0972164269940987,
+      "learning_rate": 4.555690326118961e-05,
+      "loss": 1.2773,
+      "step": 5721
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.001682765661968,
+      "learning_rate": 4.5554529561649744e-05,
+      "loss": 0.667,
+      "step": 5722
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.849434967193288,
+      "learning_rate": 4.555215529008569e-05,
+      "loss": 1.332,
+      "step": 5723
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.343934124309396,
+      "learning_rate": 4.5549780446563516e-05,
+      "loss": 1.5566,
+      "step": 5724
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8197006449519306,
+      "learning_rate": 4.554740503114933e-05,
+      "loss": 0.8145,
+      "step": 5725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7886856636618065,
+      "learning_rate": 4.554502904390922e-05,
+      "loss": 0.7241,
+      "step": 5726
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.271464281074136,
+      "learning_rate": 4.554265248490931e-05,
+      "loss": 1.6406,
+      "step": 5727
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8714659127614839,
+      "learning_rate": 4.554027535421574e-05,
+      "loss": 0.9009,
+      "step": 5728
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1594901687670323,
+      "learning_rate": 4.5537897651894684e-05,
+      "loss": 1.6582,
+      "step": 5729
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6411852753396916,
+      "learning_rate": 4.5535519378012295e-05,
+      "loss": 1.1729,
+      "step": 5730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.459631643747218,
+      "learning_rate": 4.553314053263476e-05,
+      "loss": 1.8789,
+      "step": 5731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.6197124877245646,
+      "learning_rate": 4.5530761115828274e-05,
+      "loss": 0.7949,
+      "step": 5732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0191684301066624,
+      "learning_rate": 4.552838112765907e-05,
+      "loss": 0.8999,
+      "step": 5733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.424050545999873,
+      "learning_rate": 4.552600056819338e-05,
+      "loss": 0.7725,
+      "step": 5734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.975188564726757,
+      "learning_rate": 4.552361943749744e-05,
+      "loss": 1.2217,
+      "step": 5735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.832224269209724,
+      "learning_rate": 4.552123773563753e-05,
+      "loss": 1.248,
+      "step": 5736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.590083449667202,
+      "learning_rate": 4.551885546267994e-05,
+      "loss": 1.3711,
+      "step": 5737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9214376556740878,
+      "learning_rate": 4.5516472618690944e-05,
+      "loss": 1.3184,
+      "step": 5738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9064540996460977,
+      "learning_rate": 4.551408920373688e-05,
+      "loss": 0.4419,
+      "step": 5739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.505880116401397,
+      "learning_rate": 4.551170521788405e-05,
+      "loss": 1.166,
+      "step": 5740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.942043706661267,
+      "learning_rate": 4.550932066119882e-05,
+      "loss": 1.7402,
+      "step": 5741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2850825134783643,
+      "learning_rate": 4.5506935533747544e-05,
+      "loss": 0.8433,
+      "step": 5742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.008121808046855,
+      "learning_rate": 4.55045498355966e-05,
+      "loss": 0.9131,
+      "step": 5743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.393321410371037,
+      "learning_rate": 4.550216356681238e-05,
+      "loss": 1.6348,
+      "step": 5744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0740417662186283,
+      "learning_rate": 4.54997767274613e-05,
+      "loss": 0.9434,
+      "step": 5745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.529116845346752,
+      "learning_rate": 4.5497389317609775e-05,
+      "loss": 1.5146,
+      "step": 5746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.776392199993395,
+      "learning_rate": 4.549500133732425e-05,
+      "loss": 1.5237,
+      "step": 5747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.460775876833519,
+      "learning_rate": 4.549261278667118e-05,
+      "loss": 1.3887,
+      "step": 5748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3400341841654244,
+      "learning_rate": 4.5490223665717044e-05,
+      "loss": 1.5879,
+      "step": 5749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9001492692086397,
+      "learning_rate": 4.5487833974528325e-05,
+      "loss": 0.6853,
+      "step": 5750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.32215221340704,
+      "learning_rate": 4.548544371317153e-05,
+      "loss": 0.8652,
+      "step": 5751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.632095556246375,
+      "learning_rate": 4.548305288171316e-05,
+      "loss": 0.9023,
+      "step": 5752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.801929357245416,
+      "learning_rate": 4.548066148021978e-05,
+      "loss": 1.6875,
+      "step": 5753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3674171260842383,
+      "learning_rate": 4.547826950875793e-05,
+      "loss": 0.3701,
+      "step": 5754
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.84023684182257,
+      "learning_rate": 4.547587696739417e-05,
+      "loss": 1.4131,
+      "step": 5755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.415857606944114,
+      "learning_rate": 4.547348385619509e-05,
+      "loss": 0.7798,
+      "step": 5756
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.292347292540801,
+      "learning_rate": 4.5471090175227297e-05,
+      "loss": 0.582,
+      "step": 5757
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3274602808254814,
+      "learning_rate": 4.546869592455739e-05,
+      "loss": 0.9902,
+      "step": 5758
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.639008211490824,
+      "learning_rate": 4.546630110425201e-05,
+      "loss": 1.3477,
+      "step": 5759
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.907720454778966,
+      "learning_rate": 4.54639057143778e-05,
+      "loss": 1.0469,
+      "step": 5760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9870987501925033,
+      "learning_rate": 4.5461509755001425e-05,
+      "loss": 1.0449,
+      "step": 5761
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9923705372218596,
+      "learning_rate": 4.5459113226189566e-05,
+      "loss": 0.7998,
+      "step": 5762
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.070022375168169,
+      "learning_rate": 4.5456716128008916e-05,
+      "loss": 1.2581,
+      "step": 5763
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.451759530577862,
+      "learning_rate": 4.545431846052618e-05,
+      "loss": 1.0879,
+      "step": 5764
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5004631567126503,
+      "learning_rate": 4.545192022380809e-05,
+      "loss": 1.1289,
+      "step": 5765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5089264769319946,
+      "learning_rate": 4.5449521417921384e-05,
+      "loss": 0.9314,
+      "step": 5766
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.324856724914765,
+      "learning_rate": 4.5447122042932835e-05,
+      "loss": 0.9141,
+      "step": 5767
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.648092691937035,
+      "learning_rate": 4.5444722098909185e-05,
+      "loss": 0.7793,
+      "step": 5768
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.850086713620984,
+      "learning_rate": 4.5442321585917255e-05,
+      "loss": 1.2598,
+      "step": 5769
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.2098402384968585,
+      "learning_rate": 4.543992050402383e-05,
+      "loss": 1.8145,
+      "step": 5770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6104722770257336,
+      "learning_rate": 4.543751885329574e-05,
+      "loss": 1.6719,
+      "step": 5771
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.681711599945843,
+      "learning_rate": 4.543511663379982e-05,
+      "loss": 1.8848,
+      "step": 5772
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9777329914400563,
+      "learning_rate": 4.5432713845602924e-05,
+      "loss": 1.2676,
+      "step": 5773
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.181229500994936,
+      "learning_rate": 4.5430310488771924e-05,
+      "loss": 1.0713,
+      "step": 5774
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.559118971169542,
+      "learning_rate": 4.5427906563373704e-05,
+      "loss": 1.1641,
+      "step": 5775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9212480468342965,
+      "learning_rate": 4.5425502069475146e-05,
+      "loss": 1.2698,
+      "step": 5776
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3251511883400036,
+      "learning_rate": 4.542309700714319e-05,
+      "loss": 0.4268,
+      "step": 5777
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.633637936149817,
+      "learning_rate": 4.542069137644476e-05,
+      "loss": 2.2607,
+      "step": 5778
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.711208275016207,
+      "learning_rate": 4.541828517744679e-05,
+      "loss": 0.6353,
+      "step": 5779
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6755915180917749,
+      "learning_rate": 4.5415878410216274e-05,
+      "loss": 0.6958,
+      "step": 5780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.27330651184473,
+      "learning_rate": 4.541347107482016e-05,
+      "loss": 0.605,
+      "step": 5781
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9008190773721032,
+      "learning_rate": 4.5411063171325466e-05,
+      "loss": 0.3808,
+      "step": 5782
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.831097829794126,
+      "learning_rate": 4.540865469979918e-05,
+      "loss": 1.3105,
+      "step": 5783
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.361967872556176,
+      "learning_rate": 4.540624566030836e-05,
+      "loss": 0.5347,
+      "step": 5784
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.35268214448532,
+      "learning_rate": 4.5403836052920015e-05,
+      "loss": 1.0,
+      "step": 5785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.328050082553325,
+      "learning_rate": 4.540142587770122e-05,
+      "loss": 0.7488,
+      "step": 5786
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1428492886535575,
+      "learning_rate": 4.539901513471905e-05,
+      "loss": 1.2207,
+      "step": 5787
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3407609316949403,
+      "learning_rate": 4.53966038240406e-05,
+      "loss": 1.3633,
+      "step": 5788
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.1184913430170065,
+      "learning_rate": 4.539419194573297e-05,
+      "loss": 0.5789,
+      "step": 5789
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.967010148841361,
+      "learning_rate": 4.539177949986327e-05,
+      "loss": 1.4082,
+      "step": 5790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1394010511367383,
+      "learning_rate": 4.5389366486498655e-05,
+      "loss": 1.0117,
+      "step": 5791
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.711694252400959,
+      "learning_rate": 4.538695290570627e-05,
+      "loss": 1.335,
+      "step": 5792
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6989654112977877,
+      "learning_rate": 4.5384538757553284e-05,
+      "loss": 1.584,
+      "step": 5793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.250111047335619,
+      "learning_rate": 4.538212404210688e-05,
+      "loss": 1.2168,
+      "step": 5794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7673998663782766,
+      "learning_rate": 4.537970875943427e-05,
+      "loss": 1.4883,
+      "step": 5795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.887333330901181,
+      "learning_rate": 4.5377292909602656e-05,
+      "loss": 1.709,
+      "step": 5796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3958549420814035,
+      "learning_rate": 4.5374876492679274e-05,
+      "loss": 0.7314,
+      "step": 5797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9640513933181802,
+      "learning_rate": 4.537245950873138e-05,
+      "loss": 0.8438,
+      "step": 5798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9350123948867806,
+      "learning_rate": 4.5370041957826225e-05,
+      "loss": 1.2559,
+      "step": 5799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3971202103220386,
+      "learning_rate": 4.53676238400311e-05,
+      "loss": 0.9653,
+      "step": 5800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3845764212933216,
+      "learning_rate": 4.536520515541329e-05,
+      "loss": 0.6885,
+      "step": 5801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.214767489786499,
+      "learning_rate": 4.536278590404012e-05,
+      "loss": 1.0459,
+      "step": 5802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.70206152662793,
+      "learning_rate": 4.53603660859789e-05,
+      "loss": 0.6963,
+      "step": 5803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.070718193575465,
+      "learning_rate": 4.535794570129698e-05,
+      "loss": 1.5488,
+      "step": 5804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.060657872492172,
+      "learning_rate": 4.5355524750061724e-05,
+      "loss": 1.5469,
+      "step": 5805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.474544244607017,
+      "learning_rate": 4.53531032323405e-05,
+      "loss": 1.1299,
+      "step": 5806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3725134230335483,
+      "learning_rate": 4.5350681148200693e-05,
+      "loss": 1.0127,
+      "step": 5807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.520780995598807,
+      "learning_rate": 4.534825849770973e-05,
+      "loss": 1.0925,
+      "step": 5808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.523366921208462,
+      "learning_rate": 4.5345835280935e-05,
+      "loss": 1.3398,
+      "step": 5809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.501411769535418,
+      "learning_rate": 4.5343411497943965e-05,
+      "loss": 1.1328,
+      "step": 5810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5915526308768526,
+      "learning_rate": 4.534098714880407e-05,
+      "loss": 1.3633,
+      "step": 5811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.729409008566768,
+      "learning_rate": 4.5338562233582775e-05,
+      "loss": 1.875,
+      "step": 5812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.335280485468658,
+      "learning_rate": 4.533613675234758e-05,
+      "loss": 0.8486,
+      "step": 5813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7038842561652556,
+      "learning_rate": 4.533371070516598e-05,
+      "loss": 1.3418,
+      "step": 5814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.082450210361236,
+      "learning_rate": 4.533128409210549e-05,
+      "loss": 1.5078,
+      "step": 5815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.828467890837793,
+      "learning_rate": 4.532885691323363e-05,
+      "loss": 1.5449,
+      "step": 5816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5693753471234073,
+      "learning_rate": 4.532642916861797e-05,
+      "loss": 0.9209,
+      "step": 5817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.71745655379532,
+      "learning_rate": 4.532400085832605e-05,
+      "loss": 0.957,
+      "step": 5818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6592245164693598,
+      "learning_rate": 4.532157198242546e-05,
+      "loss": 1.0947,
+      "step": 5819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4312368279813694,
+      "learning_rate": 4.531914254098381e-05,
+      "loss": 1.0332,
+      "step": 5820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7272010208586766,
+      "learning_rate": 4.5316712534068685e-05,
+      "loss": 1.2637,
+      "step": 5821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.401291873647855,
+      "learning_rate": 4.531428196174772e-05,
+      "loss": 0.7568,
+      "step": 5822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.91915006368266,
+      "learning_rate": 4.531185082408856e-05,
+      "loss": 1.1885,
+      "step": 5823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1514877583383516,
+      "learning_rate": 4.5309419121158856e-05,
+      "loss": 1.0869,
+      "step": 5824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.033388029205797,
+      "learning_rate": 4.530698685302629e-05,
+      "loss": 1.1641,
+      "step": 5825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7508378053120532,
+      "learning_rate": 4.530455401975855e-05,
+      "loss": 0.875,
+      "step": 5826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2359286644785974,
+      "learning_rate": 4.530212062142333e-05,
+      "loss": 1.3086,
+      "step": 5827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6310101687482397,
+      "learning_rate": 4.529968665808837e-05,
+      "loss": 1.0254,
+      "step": 5828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.280393885761672,
+      "learning_rate": 4.5297252129821385e-05,
+      "loss": 0.9818,
+      "step": 5829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5888401022151677,
+      "learning_rate": 4.529481703669013e-05,
+      "loss": 0.582,
+      "step": 5830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.157033057768941,
+      "learning_rate": 4.529238137876239e-05,
+      "loss": 1.0088,
+      "step": 5831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.5583916548229775,
+      "learning_rate": 4.528994515610594e-05,
+      "loss": 1.2988,
+      "step": 5832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2975415464477953,
+      "learning_rate": 4.528750836878857e-05,
+      "loss": 0.6587,
+      "step": 5833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2172594785859703,
+      "learning_rate": 4.52850710168781e-05,
+      "loss": 1.1221,
+      "step": 5834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.213567192210938,
+      "learning_rate": 4.528263310044236e-05,
+      "loss": 1.1348,
+      "step": 5835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4091479203101693,
+      "learning_rate": 4.5280194619549196e-05,
+      "loss": 1.1084,
+      "step": 5836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.697896389237292,
+      "learning_rate": 4.527775557426648e-05,
+      "loss": 0.6729,
+      "step": 5837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7271796022757373,
+      "learning_rate": 4.527531596466208e-05,
+      "loss": 0.7358,
+      "step": 5838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.442957620449095,
+      "learning_rate": 4.5272875790803895e-05,
+      "loss": 0.6304,
+      "step": 5839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.038986234848771,
+      "learning_rate": 4.527043505275982e-05,
+      "loss": 1.6758,
+      "step": 5840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.181871228265676,
+      "learning_rate": 4.5267993750597794e-05,
+      "loss": 1.7656,
+      "step": 5841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9161885812450827,
+      "learning_rate": 4.526555188438575e-05,
+      "loss": 1.0801,
+      "step": 5842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.743028145842187,
+      "learning_rate": 4.526310945419165e-05,
+      "loss": 1.7188,
+      "step": 5843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5677151528371716,
+      "learning_rate": 4.526066646008346e-05,
+      "loss": 0.3486,
+      "step": 5844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3523742694072176,
+      "learning_rate": 4.5258222902129174e-05,
+      "loss": 1.2129,
+      "step": 5845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.031249526859226,
+      "learning_rate": 4.52557787803968e-05,
+      "loss": 0.8477,
+      "step": 5846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.949263213917616,
+      "learning_rate": 4.5253334094954334e-05,
+      "loss": 0.7627,
+      "step": 5847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.365519020599159,
+      "learning_rate": 4.525088884586983e-05,
+      "loss": 1.1758,
+      "step": 5848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9651937305764282,
+      "learning_rate": 4.524844303321134e-05,
+      "loss": 0.6924,
+      "step": 5849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.671748018872453,
+      "learning_rate": 4.5245996657046916e-05,
+      "loss": 1.3008,
+      "step": 5850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.332656546487993,
+      "learning_rate": 4.524354971744465e-05,
+      "loss": 1.2324,
+      "step": 5851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.031581894570894,
+      "learning_rate": 4.5241102214472634e-05,
+      "loss": 1.2998,
+      "step": 5852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.226232350283211,
+      "learning_rate": 4.5238654148198984e-05,
+      "loss": 1.2461,
+      "step": 5853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1071111210798033,
+      "learning_rate": 4.5236205518691826e-05,
+      "loss": 1.1865,
+      "step": 5854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.58843271196138,
+      "learning_rate": 4.5233756326019316e-05,
+      "loss": 0.8867,
+      "step": 5855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.104602739881152,
+      "learning_rate": 4.5231306570249595e-05,
+      "loss": 1.3145,
+      "step": 5856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8107266794609815,
+      "learning_rate": 4.5228856251450854e-05,
+      "loss": 0.7703,
+      "step": 5857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.203067859963798,
+      "learning_rate": 4.522640536969127e-05,
+      "loss": 1.1201,
+      "step": 5858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1409791964935354,
+      "learning_rate": 4.522395392503907e-05,
+      "loss": 1.0205,
+      "step": 5859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.764807456350427,
+      "learning_rate": 4.522150191756246e-05,
+      "loss": 1.0615,
+      "step": 5860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.030582532120397,
+      "learning_rate": 4.521904934732969e-05,
+      "loss": 1.3633,
+      "step": 5861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.588604595028493,
+      "learning_rate": 4.5216596214409e-05,
+      "loss": 1.2891,
+      "step": 5862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.954681309330919,
+      "learning_rate": 4.5214142518868666e-05,
+      "loss": 1.5176,
+      "step": 5863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.067294824231496,
+      "learning_rate": 4.521168826077698e-05,
+      "loss": 0.7266,
+      "step": 5864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5002389793614643,
+      "learning_rate": 4.5209233440202234e-05,
+      "loss": 0.9707,
+      "step": 5865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5368788499185015,
+      "learning_rate": 4.5206778057212754e-05,
+      "loss": 1.6328,
+      "step": 5866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.175961211334968,
+      "learning_rate": 4.520432211187686e-05,
+      "loss": 0.7217,
+      "step": 5867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.399519738082925,
+      "learning_rate": 4.520186560426292e-05,
+      "loss": 1.1016,
+      "step": 5868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.010007221048508,
+      "learning_rate": 4.519940853443927e-05,
+      "loss": 1.4707,
+      "step": 5869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.283177563478722,
+      "learning_rate": 4.5196950902474314e-05,
+      "loss": 1.4648,
+      "step": 5870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.826752708632972,
+      "learning_rate": 4.519449270843643e-05,
+      "loss": 0.834,
+      "step": 5871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.463484385500697,
+      "learning_rate": 4.5192033952394034e-05,
+      "loss": 1.1548,
+      "step": 5872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8281260116322433,
+      "learning_rate": 4.5189574634415555e-05,
+      "loss": 0.4316,
+      "step": 5873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.476744154757164,
+      "learning_rate": 4.518711475456944e-05,
+      "loss": 1.2676,
+      "step": 5874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0878429083421848,
+      "learning_rate": 4.518465431292413e-05,
+      "loss": 1.7285,
+      "step": 5875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.250489817391399,
+      "learning_rate": 4.518219330954811e-05,
+      "loss": 0.7222,
+      "step": 5876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.6418513343757715,
+      "learning_rate": 4.517973174450987e-05,
+      "loss": 1.0107,
+      "step": 5877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8273411291468147,
+      "learning_rate": 4.517726961787791e-05,
+      "loss": 1.5527,
+      "step": 5878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3728007122195636,
+      "learning_rate": 4.517480692972075e-05,
+      "loss": 0.8823,
+      "step": 5879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.478163614259656,
+      "learning_rate": 4.517234368010692e-05,
+      "loss": 1.4375,
+      "step": 5880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4914979369361476,
+      "learning_rate": 4.5169879869104985e-05,
+      "loss": 0.6846,
+      "step": 5881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3113406600212527,
+      "learning_rate": 4.516741549678351e-05,
+      "loss": 0.9297,
+      "step": 5882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.449613645802618,
+      "learning_rate": 4.516495056321106e-05,
+      "loss": 0.6206,
+      "step": 5883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8418633207438146,
+      "learning_rate": 4.516248506845625e-05,
+      "loss": 0.9258,
+      "step": 5884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0573816078581006,
+      "learning_rate": 4.516001901258768e-05,
+      "loss": 1.6094,
+      "step": 5885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4282609977587013,
+      "learning_rate": 4.5157552395674e-05,
+      "loss": 1.0586,
+      "step": 5886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.301589432537485,
+      "learning_rate": 4.515508521778384e-05,
+      "loss": 0.7617,
+      "step": 5887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9962255257445234,
+      "learning_rate": 4.515261747898586e-05,
+      "loss": 1.293,
+      "step": 5888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.703114569926965,
+      "learning_rate": 4.5150149179348744e-05,
+      "loss": 0.625,
+      "step": 5889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.364068731079851,
+      "learning_rate": 4.5147680318941176e-05,
+      "loss": 0.7402,
+      "step": 5890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.1680597252598135,
+      "learning_rate": 4.514521089783187e-05,
+      "loss": 1.1699,
+      "step": 5891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.65202965653352,
+      "learning_rate": 4.5142740916089544e-05,
+      "loss": 1.2012,
+      "step": 5892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.863864396939495,
+      "learning_rate": 4.514027037378294e-05,
+      "loss": 0.894,
+      "step": 5893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.092226366104357,
+      "learning_rate": 4.5137799270980806e-05,
+      "loss": 1.0938,
+      "step": 5894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2756462772153117,
+      "learning_rate": 4.513532760775192e-05,
+      "loss": 0.9971,
+      "step": 5895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.757118312260146,
+      "learning_rate": 4.513285538416507e-05,
+      "loss": 1.334,
+      "step": 5896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0812840364439555,
+      "learning_rate": 4.5130382600289046e-05,
+      "loss": 1.1406,
+      "step": 5897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.977772324612068,
+      "learning_rate": 4.512790925619266e-05,
+      "loss": 1.2051,
+      "step": 5898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7742035677422783,
+      "learning_rate": 4.512543535194477e-05,
+      "loss": 0.976,
+      "step": 5899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8936715232803385,
+      "learning_rate": 4.512296088761419e-05,
+      "loss": 0.3728,
+      "step": 5900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.525849314703789,
+      "learning_rate": 4.512048586326981e-05,
+      "loss": 1.1445,
+      "step": 5901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.275876549173447,
+      "learning_rate": 4.51180102789805e-05,
+      "loss": 0.4797,
+      "step": 5902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1439649747300145,
+      "learning_rate": 4.511553413481515e-05,
+      "loss": 1.25,
+      "step": 5903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.953613854137679,
+      "learning_rate": 4.5113057430842674e-05,
+      "loss": 0.9268,
+      "step": 5904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.561032042978637,
+      "learning_rate": 4.5110580167132e-05,
+      "loss": 0.8086,
+      "step": 5905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.177668357498507,
+      "learning_rate": 4.510810234375207e-05,
+      "loss": 0.8896,
+      "step": 5906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.26092176008607,
+      "learning_rate": 4.5105623960771834e-05,
+      "loss": 0.9688,
+      "step": 5907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.956361886261233,
+      "learning_rate": 4.510314501826027e-05,
+      "loss": 1.1377,
+      "step": 5908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.099734621082719,
+      "learning_rate": 4.5100665516286366e-05,
+      "loss": 0.9355,
+      "step": 5909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.113490406340391,
+      "learning_rate": 4.5098185454919116e-05,
+      "loss": 0.8184,
+      "step": 5910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.160519588094455,
+      "learning_rate": 4.509570483422755e-05,
+      "loss": 0.5228,
+      "step": 5911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.7190450613979,
+      "learning_rate": 4.50932236542807e-05,
+      "loss": 1.1348,
+      "step": 5912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1726686584034622,
+      "learning_rate": 4.509074191514762e-05,
+      "loss": 0.9756,
+      "step": 5913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.6905458854087545,
+      "learning_rate": 4.508825961689737e-05,
+      "loss": 1.1426,
+      "step": 5914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.106767121161079,
+      "learning_rate": 4.508577675959903e-05,
+      "loss": 1.8535,
+      "step": 5915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.264002054475624,
+      "learning_rate": 4.50832933433217e-05,
+      "loss": 0.9736,
+      "step": 5916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3929671239894086,
+      "learning_rate": 4.508080936813449e-05,
+      "loss": 1.5117,
+      "step": 5917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.871602159190852,
+      "learning_rate": 4.507832483410654e-05,
+      "loss": 1.8662,
+      "step": 5918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.075975615435777,
+      "learning_rate": 4.507583974130697e-05,
+      "loss": 1.8984,
+      "step": 5919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.286417939686016,
+      "learning_rate": 4.507335408980496e-05,
+      "loss": 0.6445,
+      "step": 5920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6946199716067265,
+      "learning_rate": 4.507086787966968e-05,
+      "loss": 1.3076,
+      "step": 5921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4212196694021797,
+      "learning_rate": 4.506838111097031e-05,
+      "loss": 1.1494,
+      "step": 5922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.306554002444465,
+      "learning_rate": 4.506589378377608e-05,
+      "loss": 1.3223,
+      "step": 5923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.434291096760679,
+      "learning_rate": 4.5063405898156176e-05,
+      "loss": 0.9023,
+      "step": 5924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0866718432532037,
+      "learning_rate": 4.506091745417986e-05,
+      "loss": 1.4219,
+      "step": 5925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7480291760490814,
+      "learning_rate": 4.5058428451916376e-05,
+      "loss": 1.0371,
+      "step": 5926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.292325555119315,
+      "learning_rate": 4.5055938891435e-05,
+      "loss": 1.0,
+      "step": 5927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.277258424812636,
+      "learning_rate": 4.5053448772805007e-05,
+      "loss": 1.3574,
+      "step": 5928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.988771036668132,
+      "learning_rate": 4.5050958096095696e-05,
+      "loss": 1.4102,
+      "step": 5929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.837882739951184,
+      "learning_rate": 4.5048466861376394e-05,
+      "loss": 0.7319,
+      "step": 5930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4493236763397133,
+      "learning_rate": 4.504597506871641e-05,
+      "loss": 1.2129,
+      "step": 5931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6600702078595946,
+      "learning_rate": 4.50434827181851e-05,
+      "loss": 0.9844,
+      "step": 5932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.367734537400595,
+      "learning_rate": 4.504098980985183e-05,
+      "loss": 0.9072,
+      "step": 5933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1473789754783588,
+      "learning_rate": 4.5038496343785975e-05,
+      "loss": 1.3467,
+      "step": 5934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8378445857199783,
+      "learning_rate": 4.5036002320056914e-05,
+      "loss": 1.3379,
+      "step": 5935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.954380821851276,
+      "learning_rate": 4.5033507738734073e-05,
+      "loss": 0.4834,
+      "step": 5936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.648725285831097,
+      "learning_rate": 4.503101259988687e-05,
+      "loss": 1.2842,
+      "step": 5937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.561180539643771,
+      "learning_rate": 4.502851690358474e-05,
+      "loss": 0.8906,
+      "step": 5938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3552008667351254,
+      "learning_rate": 4.5026020649897135e-05,
+      "loss": 1.0039,
+      "step": 5939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7777519679513938,
+      "learning_rate": 4.502352383889352e-05,
+      "loss": 0.8285,
+      "step": 5940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.500654395727904,
+      "learning_rate": 4.50210264706434e-05,
+      "loss": 0.7556,
+      "step": 5941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.813633574705112,
+      "learning_rate": 4.501852854521626e-05,
+      "loss": 1.0596,
+      "step": 5942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3200480995370754,
+      "learning_rate": 4.5016030062681617e-05,
+      "loss": 0.6641,
+      "step": 5943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.159279757157377,
+      "learning_rate": 4.5013531023109014e-05,
+      "loss": 1.4297,
+      "step": 5944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.573192432364092,
+      "learning_rate": 4.5011031426567986e-05,
+      "loss": 1.2354,
+      "step": 5945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.323882531054547,
+      "learning_rate": 4.50085312731281e-05,
+      "loss": 0.5344,
+      "step": 5946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.742496048141199,
+      "learning_rate": 4.500603056285893e-05,
+      "loss": 1.1282,
+      "step": 5947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.324572008999595,
+      "learning_rate": 4.5003529295830075e-05,
+      "loss": 1.3164,
+      "step": 5948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.267370143818514,
+      "learning_rate": 4.5001027472111146e-05,
+      "loss": 1.1602,
+      "step": 5949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.994529493867058,
+      "learning_rate": 4.499852509177177e-05,
+      "loss": 1.8262,
+      "step": 5950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.546654581337992,
+      "learning_rate": 4.4996022154881575e-05,
+      "loss": 0.8247,
+      "step": 5951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.3882392021502294,
+      "learning_rate": 4.4993518661510226e-05,
+      "loss": 1.1201,
+      "step": 5952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.965507744580429,
+      "learning_rate": 4.4991014611727386e-05,
+      "loss": 1.2793,
+      "step": 5953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.945095543750021,
+      "learning_rate": 4.4988510005602756e-05,
+      "loss": 1.5762,
+      "step": 5954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.367787110139636,
+      "learning_rate": 4.498600484320602e-05,
+      "loss": 1.1055,
+      "step": 5955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.557459388797937,
+      "learning_rate": 4.498349912460692e-05,
+      "loss": 1.4653,
+      "step": 5956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6169000723990434,
+      "learning_rate": 4.498099284987516e-05,
+      "loss": 1.4043,
+      "step": 5957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.744978408683958,
+      "learning_rate": 4.497848601908051e-05,
+      "loss": 0.6221,
+      "step": 5958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0880056365957795,
+      "learning_rate": 4.497597863229273e-05,
+      "loss": 1.5117,
+      "step": 5959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.3852197356719715,
+      "learning_rate": 4.49734706895816e-05,
+      "loss": 0.9268,
+      "step": 5960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.493771587208559,
+      "learning_rate": 4.49709621910169e-05,
+      "loss": 1.248,
+      "step": 5961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.039010082620464,
+      "learning_rate": 4.496845313666847e-05,
+      "loss": 1.249,
+      "step": 5962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.338574076970716,
+      "learning_rate": 4.4965943526606096e-05,
+      "loss": 1.5996,
+      "step": 5963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1215027472779204,
+      "learning_rate": 4.496343336089965e-05,
+      "loss": 1.1589,
+      "step": 5964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.209734776458398,
+      "learning_rate": 4.4960922639618985e-05,
+      "loss": 0.8799,
+      "step": 5965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.068093966252066,
+      "learning_rate": 4.495841136283397e-05,
+      "loss": 1.7812,
+      "step": 5966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.709836620881661,
+      "learning_rate": 4.4955899530614484e-05,
+      "loss": 1.1426,
+      "step": 5967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.901656830260896,
+      "learning_rate": 4.495338714303045e-05,
+      "loss": 2.2969,
+      "step": 5968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5467571395095416,
+      "learning_rate": 4.495087420015177e-05,
+      "loss": 0.9551,
+      "step": 5969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5144989244733753,
+      "learning_rate": 4.494836070204838e-05,
+      "loss": 1.4141,
+      "step": 5970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.722018048078183,
+      "learning_rate": 4.494584664879024e-05,
+      "loss": 1.2441,
+      "step": 5971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6341002623086562,
+      "learning_rate": 4.4943332040447295e-05,
+      "loss": 0.9395,
+      "step": 5972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.316340293933221,
+      "learning_rate": 4.4940816877089544e-05,
+      "loss": 0.7373,
+      "step": 5973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0629296288814705,
+      "learning_rate": 4.493830115878698e-05,
+      "loss": 0.6015,
+      "step": 5974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.91898941920507,
+      "learning_rate": 4.493578488560961e-05,
+      "loss": 1.501,
+      "step": 5975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.82272754806245,
+      "learning_rate": 4.4933268057627464e-05,
+      "loss": 0.6851,
+      "step": 5976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.700389646742147,
+      "learning_rate": 4.4930750674910584e-05,
+      "loss": 0.9561,
+      "step": 5977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.02837654655752,
+      "learning_rate": 4.492823273752902e-05,
+      "loss": 1.5625,
+      "step": 5978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2711745072931153,
+      "learning_rate": 4.492571424555285e-05,
+      "loss": 1.0645,
+      "step": 5979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.256874602451675,
+      "learning_rate": 4.492319519905217e-05,
+      "loss": 1.6016,
+      "step": 5980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3851139148562317,
+      "learning_rate": 4.492067559809708e-05,
+      "loss": 0.958,
+      "step": 5981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.998448288477204,
+      "learning_rate": 4.49181554427577e-05,
+      "loss": 1.0273,
+      "step": 5982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.302823454085892,
+      "learning_rate": 4.491563473310416e-05,
+      "loss": 0.4585,
+      "step": 5983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.46584053710815,
+      "learning_rate": 4.491311346920661e-05,
+      "loss": 0.8591,
+      "step": 5984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.2954497543831245,
+      "learning_rate": 4.491059165113522e-05,
+      "loss": 0.9248,
+      "step": 5985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.2997277095019895,
+      "learning_rate": 4.490806927896017e-05,
+      "loss": 0.8447,
+      "step": 5986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1055288783955373,
+      "learning_rate": 4.490554635275166e-05,
+      "loss": 0.7744,
+      "step": 5987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7943939993505873,
+      "learning_rate": 4.4903022872579895e-05,
+      "loss": 0.709,
+      "step": 5988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.960101731846123,
+      "learning_rate": 4.4900498838515105e-05,
+      "loss": 1.1073,
+      "step": 5989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0840319967428345,
+      "learning_rate": 4.4897974250627533e-05,
+      "loss": 0.8682,
+      "step": 5990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.232361835025859,
+      "learning_rate": 4.489544910898744e-05,
+      "loss": 0.8863,
+      "step": 5991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5949897286343973,
+      "learning_rate": 4.48929234136651e-05,
+      "loss": 0.7998,
+      "step": 5992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.015931425190267,
+      "learning_rate": 4.489039716473079e-05,
+      "loss": 1.0312,
+      "step": 5993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2341659554813917,
+      "learning_rate": 4.4887870362254835e-05,
+      "loss": 1.4766,
+      "step": 5994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5231732209411932,
+      "learning_rate": 4.488534300630754e-05,
+      "loss": 0.9482,
+      "step": 5995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2075567379460987,
+      "learning_rate": 4.488281509695925e-05,
+      "loss": 0.9089,
+      "step": 5996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.632116118116149,
+      "learning_rate": 4.488028663428031e-05,
+      "loss": 0.7754,
+      "step": 5997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1792306782755597,
+      "learning_rate": 4.487775761834107e-05,
+      "loss": 1.2559,
+      "step": 5998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6391689642689515,
+      "learning_rate": 4.487522804921195e-05,
+      "loss": 1.668,
+      "step": 5999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.177228737193237,
+      "learning_rate": 4.487269792696331e-05,
+      "loss": 0.7129,
+      "step": 6000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.288405392432098,
+      "learning_rate": 4.4870167251665583e-05,
+      "loss": 1.2461,
+      "step": 6001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.778155353816478,
+      "learning_rate": 4.486763602338919e-05,
+      "loss": 1.3379,
+      "step": 6002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.073876678471782,
+      "learning_rate": 4.486510424220457e-05,
+      "loss": 1.1787,
+      "step": 6003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.512585812142325,
+      "learning_rate": 4.4862571908182194e-05,
+      "loss": 1.4922,
+      "step": 6004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3403743924668303,
+      "learning_rate": 4.486003902139252e-05,
+      "loss": 0.9248,
+      "step": 6005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.213521680554687,
+      "learning_rate": 4.4857505581906054e-05,
+      "loss": 1.0264,
+      "step": 6006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.279544976990808,
+      "learning_rate": 4.485497158979328e-05,
+      "loss": 1.1172,
+      "step": 6007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.106508468933249,
+      "learning_rate": 4.485243704512474e-05,
+      "loss": 0.9668,
+      "step": 6008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.553478990145479,
+      "learning_rate": 4.484990194797095e-05,
+      "loss": 0.7598,
+      "step": 6009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2605384289926675,
+      "learning_rate": 4.484736629840248e-05,
+      "loss": 0.4001,
+      "step": 6010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.236097729860044,
+      "learning_rate": 4.484483009648988e-05,
+      "loss": 1.7188,
+      "step": 6011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6223149056333757,
+      "learning_rate": 4.484229334230373e-05,
+      "loss": 1.0166,
+      "step": 6012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5034863957606293,
+      "learning_rate": 4.483975603591465e-05,
+      "loss": 0.8008,
+      "step": 6013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.986816227042296,
+      "learning_rate": 4.483721817739322e-05,
+      "loss": 1.0332,
+      "step": 6014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.959364666827879,
+      "learning_rate": 4.483467976681009e-05,
+      "loss": 0.8296,
+      "step": 6015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1637441969524023,
+      "learning_rate": 4.48321408042359e-05,
+      "loss": 1.4512,
+      "step": 6016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.882136237159996,
+      "learning_rate": 4.4829601289741306e-05,
+      "loss": 1.0674,
+      "step": 6017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3719522094222216,
+      "learning_rate": 4.482706122339698e-05,
+      "loss": 0.5439,
+      "step": 6018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.541119790471951,
+      "learning_rate": 4.4824520605273615e-05,
+      "loss": 0.8782,
+      "step": 6019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.087491717293577,
+      "learning_rate": 4.48219794354419e-05,
+      "loss": 1.3076,
+      "step": 6020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.311405203713713,
+      "learning_rate": 4.481943771397257e-05,
+      "loss": 1.2061,
+      "step": 6021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.755985778685384,
+      "learning_rate": 4.481689544093637e-05,
+      "loss": 1.0967,
+      "step": 6022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.865228883078837,
+      "learning_rate": 4.481435261640402e-05,
+      "loss": 0.8125,
+      "step": 6023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5281239739452492,
+      "learning_rate": 4.481180924044631e-05,
+      "loss": 1.7559,
+      "step": 6024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.524325754771922,
+      "learning_rate": 4.480926531313402e-05,
+      "loss": 1.3691,
+      "step": 6025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.444778335290055,
+      "learning_rate": 4.480672083453794e-05,
+      "loss": 0.9336,
+      "step": 6026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0866086644448636,
+      "learning_rate": 4.4804175804728875e-05,
+      "loss": 1.3262,
+      "step": 6027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.429276017551712,
+      "learning_rate": 4.4801630223777665e-05,
+      "loss": 1.2695,
+      "step": 6028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.5712427590374265,
+      "learning_rate": 4.479908409175515e-05,
+      "loss": 1.0723,
+      "step": 6029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6200647972240463,
+      "learning_rate": 4.4796537408732176e-05,
+      "loss": 1.437,
+      "step": 6030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8357303428126657,
+      "learning_rate": 4.479399017477963e-05,
+      "loss": 1.4668,
+      "step": 6031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.209082755638333,
+      "learning_rate": 4.479144238996839e-05,
+      "loss": 1.0557,
+      "step": 6032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.20551147878731,
+      "learning_rate": 4.478889405436937e-05,
+      "loss": 1.3965,
+      "step": 6033
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9123293757177455,
+      "learning_rate": 4.478634516805349e-05,
+      "loss": 1.1016,
+      "step": 6034
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9106726681060127,
+      "learning_rate": 4.478379573109167e-05,
+      "loss": 0.7466,
+      "step": 6035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.193908570158467,
+      "learning_rate": 4.478124574355487e-05,
+      "loss": 1.2012,
+      "step": 6036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.954020989607084,
+      "learning_rate": 4.477869520551406e-05,
+      "loss": 1.7695,
+      "step": 6037
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5930012001243723,
+      "learning_rate": 4.477614411704021e-05,
+      "loss": 1.1543,
+      "step": 6038
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.428480332934057,
+      "learning_rate": 4.477359247820432e-05,
+      "loss": 1.3477,
+      "step": 6039
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.413836621460378,
+      "learning_rate": 4.47710402890774e-05,
+      "loss": 1.9531,
+      "step": 6040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.659083515549637,
+      "learning_rate": 4.476848754973049e-05,
+      "loss": 0.8535,
+      "step": 6041
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.098366501687165,
+      "learning_rate": 4.476593426023461e-05,
+      "loss": 1.459,
+      "step": 6042
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3102599971155757,
+      "learning_rate": 4.476338042066082e-05,
+      "loss": 1.4434,
+      "step": 6043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.2210035985124685,
+      "learning_rate": 4.4760826031080205e-05,
+      "loss": 1.5039,
+      "step": 6044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7847026357140858,
+      "learning_rate": 4.4758271091563844e-05,
+      "loss": 1.2578,
+      "step": 6045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7724924115970597,
+      "learning_rate": 4.475571560218284e-05,
+      "loss": 1.1416,
+      "step": 6046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.447682822942472,
+      "learning_rate": 4.475315956300832e-05,
+      "loss": 1.5938,
+      "step": 6047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8273408952538224,
+      "learning_rate": 4.4750602974111407e-05,
+      "loss": 0.8174,
+      "step": 6048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.298877438837061,
+      "learning_rate": 4.474804583556325e-05,
+      "loss": 1.6074,
+      "step": 6049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.909792095227734,
+      "learning_rate": 4.474548814743502e-05,
+      "loss": 0.7031,
+      "step": 6050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.067081968118422,
+      "learning_rate": 4.47429299097979e-05,
+      "loss": 1.0127,
+      "step": 6051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5490672860730914,
+      "learning_rate": 4.474037112272307e-05,
+      "loss": 1.3496,
+      "step": 6052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.631663544787576,
+      "learning_rate": 4.4737811786281746e-05,
+      "loss": 1.2207,
+      "step": 6053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.640736154860152,
+      "learning_rate": 4.473525190054516e-05,
+      "loss": 1.5029,
+      "step": 6054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.556052034871099,
+      "learning_rate": 4.4732691465584556e-05,
+      "loss": 2.0859,
+      "step": 6055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.713735142776535,
+      "learning_rate": 4.473013048147117e-05,
+      "loss": 1.252,
+      "step": 6056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.006462122498311,
+      "learning_rate": 4.472756894827629e-05,
+      "loss": 1.0088,
+      "step": 6057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.904257983467856,
+      "learning_rate": 4.472500686607119e-05,
+      "loss": 1.0762,
+      "step": 6058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.494178124557946,
+      "learning_rate": 4.472244423492719e-05,
+      "loss": 0.7124,
+      "step": 6059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.182192557711475,
+      "learning_rate": 4.471988105491558e-05,
+      "loss": 1.0508,
+      "step": 6060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.557636473724644,
+      "learning_rate": 4.4717317326107725e-05,
+      "loss": 0.6863,
+      "step": 6061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.913724999899139,
+      "learning_rate": 4.4714753048574946e-05,
+      "loss": 1.4824,
+      "step": 6062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.438553166555303,
+      "learning_rate": 4.471218822238862e-05,
+      "loss": 1.1533,
+      "step": 6063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.706550227015276,
+      "learning_rate": 4.4709622847620114e-05,
+      "loss": 1.623,
+      "step": 6064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.500772799596822,
+      "learning_rate": 4.470705692434084e-05,
+      "loss": 0.8896,
+      "step": 6065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4649754513177813,
+      "learning_rate": 4.4704490452622184e-05,
+      "loss": 1.4785,
+      "step": 6066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9407596764042507,
+      "learning_rate": 4.470192343253559e-05,
+      "loss": 1.2451,
+      "step": 6067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.033138512309379,
+      "learning_rate": 4.469935586415248e-05,
+      "loss": 1.0117,
+      "step": 6068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.313324449800969,
+      "learning_rate": 4.469678774754431e-05,
+      "loss": 0.8401,
+      "step": 6069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3608096983759195,
+      "learning_rate": 4.4694219082782565e-05,
+      "loss": 1.1831,
+      "step": 6070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.714388706598898,
+      "learning_rate": 4.4691649869938715e-05,
+      "loss": 0.7861,
+      "step": 6071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.402120616711026,
+      "learning_rate": 4.468908010908427e-05,
+      "loss": 1.4014,
+      "step": 6072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.307716100590514,
+      "learning_rate": 4.4686509800290746e-05,
+      "loss": 0.8027,
+      "step": 6073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.286869776541645,
+      "learning_rate": 4.468393894362965e-05,
+      "loss": 1.3594,
+      "step": 6074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7171641142037863,
+      "learning_rate": 4.4681367539172567e-05,
+      "loss": 0.9612,
+      "step": 6075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.818589160575229,
+      "learning_rate": 4.467879558699103e-05,
+      "loss": 1.0508,
+      "step": 6076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8567504511163753,
+      "learning_rate": 4.467622308715662e-05,
+      "loss": 0.6953,
+      "step": 6077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.692504848510617,
+      "learning_rate": 4.467365003974094e-05,
+      "loss": 0.9121,
+      "step": 6078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.966080643838556,
+      "learning_rate": 4.467107644481558e-05,
+      "loss": 1.9707,
+      "step": 6079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9407739357917586,
+      "learning_rate": 4.4668502302452175e-05,
+      "loss": 0.7534,
+      "step": 6080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.621881721860796,
+      "learning_rate": 4.466592761272236e-05,
+      "loss": 0.8232,
+      "step": 6081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6421045074873755,
+      "learning_rate": 4.4663352375697776e-05,
+      "loss": 1.1489,
+      "step": 6082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7317829154068125,
+      "learning_rate": 4.4660776591450105e-05,
+      "loss": 0.8281,
+      "step": 6083
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.731041582129707,
+      "learning_rate": 4.465820026005103e-05,
+      "loss": 1.0879,
+      "step": 6084
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.218199558125194,
+      "learning_rate": 4.4655623381572245e-05,
+      "loss": 0.9109,
+      "step": 6085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1657558101578744,
+      "learning_rate": 4.465304595608546e-05,
+      "loss": 1.3936,
+      "step": 6086
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1034665469480704,
+      "learning_rate": 4.46504679836624e-05,
+      "loss": 0.5759,
+      "step": 6087
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8857606942088463,
+      "learning_rate": 4.464788946437483e-05,
+      "loss": 1.0508,
+      "step": 6088
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4989752576625133,
+      "learning_rate": 4.464531039829448e-05,
+      "loss": 0.4432,
+      "step": 6089
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9167523689394455,
+      "learning_rate": 4.4642730785493145e-05,
+      "loss": 0.8555,
+      "step": 6090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9870596589093745,
+      "learning_rate": 4.464015062604262e-05,
+      "loss": 1.1084,
+      "step": 6091
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0920649338048425,
+      "learning_rate": 4.4637569920014684e-05,
+      "loss": 1.3848,
+      "step": 6092
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.399284029355512,
+      "learning_rate": 4.4634988667481176e-05,
+      "loss": 1.0254,
+      "step": 6093
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.508753864760491,
+      "learning_rate": 4.4632406868513926e-05,
+      "loss": 1.1777,
+      "step": 6094
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.313742206784484,
+      "learning_rate": 4.4629824523184784e-05,
+      "loss": 1.1709,
+      "step": 6095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.001341594778615,
+      "learning_rate": 4.462724163156562e-05,
+      "loss": 0.3887,
+      "step": 6096
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2716681552843836,
+      "learning_rate": 4.4624658193728305e-05,
+      "loss": 1.3906,
+      "step": 6097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.130331602964619,
+      "learning_rate": 4.4622074209744755e-05,
+      "loss": 1.2402,
+      "step": 6098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.597795100063935,
+      "learning_rate": 4.461948967968686e-05,
+      "loss": 0.8279,
+      "step": 6099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3840467914559307,
+      "learning_rate": 4.461690460362655e-05,
+      "loss": 1.0547,
+      "step": 6100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.303912533570081,
+      "learning_rate": 4.461431898163578e-05,
+      "loss": 1.2031,
+      "step": 6101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7554614714571524,
+      "learning_rate": 4.461173281378649e-05,
+      "loss": 0.9629,
+      "step": 6102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.465631452736098,
+      "learning_rate": 4.460914610015067e-05,
+      "loss": 1.6719,
+      "step": 6103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.092969281246983,
+      "learning_rate": 4.460655884080028e-05,
+      "loss": 0.9138,
+      "step": 6104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4466739070102648,
+      "learning_rate": 4.460397103580735e-05,
+      "loss": 0.8428,
+      "step": 6105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.7682624498183515,
+      "learning_rate": 4.460138268524389e-05,
+      "loss": 1.4846,
+      "step": 6106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.715649325015438,
+      "learning_rate": 4.459879378918192e-05,
+      "loss": 1.2803,
+      "step": 6107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 30.78869873389524,
+      "learning_rate": 4.459620434769351e-05,
+      "loss": 1.8828,
+      "step": 6108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.341805936081563,
+      "learning_rate": 4.45936143608507e-05,
+      "loss": 1.5488,
+      "step": 6109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.875458058894602,
+      "learning_rate": 4.459102382872559e-05,
+      "loss": 1.1084,
+      "step": 6110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0013453328070554,
+      "learning_rate": 4.458843275139025e-05,
+      "loss": 0.9438,
+      "step": 6111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.836610256870379,
+      "learning_rate": 4.45858411289168e-05,
+      "loss": 0.432,
+      "step": 6112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.167773978617094,
+      "learning_rate": 4.4583248961377376e-05,
+      "loss": 0.9619,
+      "step": 6113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7743981645256843,
+      "learning_rate": 4.45806562488441e-05,
+      "loss": 0.7104,
+      "step": 6114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.494976003245612,
+      "learning_rate": 4.4578062991389136e-05,
+      "loss": 1.3691,
+      "step": 6115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4331119850373057,
+      "learning_rate": 4.457546918908464e-05,
+      "loss": 1.043,
+      "step": 6116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6109554050454844,
+      "learning_rate": 4.457287484200281e-05,
+      "loss": 0.9922,
+      "step": 6117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4012544729924676,
+      "learning_rate": 4.457027995021584e-05,
+      "loss": 0.688,
+      "step": 6118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1045434880945404,
+      "learning_rate": 4.4567684513795955e-05,
+      "loss": 1.2637,
+      "step": 6119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.964864556018377,
+      "learning_rate": 4.456508853281537e-05,
+      "loss": 1.1914,
+      "step": 6120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7948302740408373,
+      "learning_rate": 4.4562492007346334e-05,
+      "loss": 0.3628,
+      "step": 6121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.204491210593266,
+      "learning_rate": 4.455989493746111e-05,
+      "loss": 1.4316,
+      "step": 6122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.019908941286544,
+      "learning_rate": 4.4557297323231974e-05,
+      "loss": 1.1885,
+      "step": 6123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8614574416170417,
+      "learning_rate": 4.455469916473122e-05,
+      "loss": 0.8203,
+      "step": 6124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.589524285361135,
+      "learning_rate": 4.4552100462031146e-05,
+      "loss": 0.5044,
+      "step": 6125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6083003834783027,
+      "learning_rate": 4.454950121520407e-05,
+      "loss": 2.1953,
+      "step": 6126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6674372163360045,
+      "learning_rate": 4.4546901424322335e-05,
+      "loss": 0.5688,
+      "step": 6127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5301698805576676,
+      "learning_rate": 4.4544301089458296e-05,
+      "loss": 1.623,
+      "step": 6128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.290469632008465,
+      "learning_rate": 4.4541700210684315e-05,
+      "loss": 1.0786,
+      "step": 6129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.424442121576497,
+      "learning_rate": 4.453909878807276e-05,
+      "loss": 0.9951,
+      "step": 6130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0327738987675787,
+      "learning_rate": 4.453649682169605e-05,
+      "loss": 0.8701,
+      "step": 6131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.019639375634906,
+      "learning_rate": 4.453389431162659e-05,
+      "loss": 1.7656,
+      "step": 6132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.113214481464454,
+      "learning_rate": 4.45312912579368e-05,
+      "loss": 1.2451,
+      "step": 6133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9789946899651238,
+      "learning_rate": 4.452868766069912e-05,
+      "loss": 0.7012,
+      "step": 6134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.152253080050632,
+      "learning_rate": 4.452608351998602e-05,
+      "loss": 1.7031,
+      "step": 6135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.804883763479157,
+      "learning_rate": 4.452347883586996e-05,
+      "loss": 1.5078,
+      "step": 6136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.704536025611522,
+      "learning_rate": 4.4520873608423436e-05,
+      "loss": 0.7036,
+      "step": 6137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.119291049910658,
+      "learning_rate": 4.451826783771894e-05,
+      "loss": 0.6934,
+      "step": 6138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.586465054634963,
+      "learning_rate": 4.4515661523829005e-05,
+      "loss": 0.751,
+      "step": 6139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.66721665154713,
+      "learning_rate": 4.4513054666826146e-05,
+      "loss": 1.3643,
+      "step": 6140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.536279931867093,
+      "learning_rate": 4.451044726678293e-05,
+      "loss": 1.6191,
+      "step": 6141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2423453051298767,
+      "learning_rate": 4.450783932377191e-05,
+      "loss": 0.8994,
+      "step": 6142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9728971727066686,
+      "learning_rate": 4.450523083786565e-05,
+      "loss": 1.0547,
+      "step": 6143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8069244639080204,
+      "learning_rate": 4.4502621809136755e-05,
+      "loss": 1.084,
+      "step": 6144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.413946477664906,
+      "learning_rate": 4.450001223765785e-05,
+      "loss": 1.5156,
+      "step": 6145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.705412596716033,
+      "learning_rate": 4.449740212350153e-05,
+      "loss": 0.8662,
+      "step": 6146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4293540406850287,
+      "learning_rate": 4.449479146674045e-05,
+      "loss": 1.5234,
+      "step": 6147
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.817031854359361,
+      "learning_rate": 4.4492180267447264e-05,
+      "loss": 1.3242,
+      "step": 6148
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.281055844330188,
+      "learning_rate": 4.4489568525694635e-05,
+      "loss": 1.5117,
+      "step": 6149
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.316874182096536,
+      "learning_rate": 4.4486956241555246e-05,
+      "loss": 1.4746,
+      "step": 6150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9108412436870414,
+      "learning_rate": 4.44843434151018e-05,
+      "loss": 0.6323,
+      "step": 6151
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5351786407267434,
+      "learning_rate": 4.4481730046407004e-05,
+      "loss": 1.4248,
+      "step": 6152
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8028150019941926,
+      "learning_rate": 4.44791161355436e-05,
+      "loss": 0.8721,
+      "step": 6153
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.090733989161184,
+      "learning_rate": 4.4476501682584316e-05,
+      "loss": 1.1973,
+      "step": 6154
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.60566887496718,
+      "learning_rate": 4.4473886687601915e-05,
+      "loss": 1.0332,
+      "step": 6155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.702576673127208,
+      "learning_rate": 4.4471271150669194e-05,
+      "loss": 1.1055,
+      "step": 6156
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.081801821453259,
+      "learning_rate": 4.4468655071858914e-05,
+      "loss": 0.973,
+      "step": 6157
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5140693443981394,
+      "learning_rate": 4.446603845124388e-05,
+      "loss": 0.7349,
+      "step": 6158
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.03719118270389,
+      "learning_rate": 4.446342128889693e-05,
+      "loss": 1.293,
+      "step": 6159
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.997684762393435,
+      "learning_rate": 4.446080358489089e-05,
+      "loss": 0.9043,
+      "step": 6160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.68393200392703,
+      "learning_rate": 4.44581853392986e-05,
+      "loss": 1.5986,
+      "step": 6161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9021194802630315,
+      "learning_rate": 4.445556655219294e-05,
+      "loss": 0.7109,
+      "step": 6162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.23624603254074,
+      "learning_rate": 4.445294722364678e-05,
+      "loss": 1.3887,
+      "step": 6163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.61653495801051,
+      "learning_rate": 4.4450327353733025e-05,
+      "loss": 1.0508,
+      "step": 6164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.681709135318159,
+      "learning_rate": 4.444770694252457e-05,
+      "loss": 0.9688,
+      "step": 6165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5872430689449186,
+      "learning_rate": 4.444508599009436e-05,
+      "loss": 1.4727,
+      "step": 6166
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5393977017318567,
+      "learning_rate": 4.4442464496515306e-05,
+      "loss": 1.4395,
+      "step": 6167
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.6946096798761365,
+      "learning_rate": 4.443984246186039e-05,
+      "loss": 1.2695,
+      "step": 6168
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6633508352564776,
+      "learning_rate": 4.443721988620257e-05,
+      "loss": 0.7881,
+      "step": 6169
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8960139998102856,
+      "learning_rate": 4.443459676961483e-05,
+      "loss": 1.3182,
+      "step": 6170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.081462619092396,
+      "learning_rate": 4.443197311217018e-05,
+      "loss": 0.9937,
+      "step": 6171
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.326594470915531,
+      "learning_rate": 4.442934891394163e-05,
+      "loss": 1.6504,
+      "step": 6172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.349508002703299,
+      "learning_rate": 4.44267241750022e-05,
+      "loss": 1.0381,
+      "step": 6173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.682632029644735,
+      "learning_rate": 4.442409889542496e-05,
+      "loss": 1.5996,
+      "step": 6174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5484653492428544,
+      "learning_rate": 4.442147307528294e-05,
+      "loss": 1.3057,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.190568397589168,
+      "learning_rate": 4.441884671464924e-05,
+      "loss": 1.1992,
+      "step": 6176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2905167786975222,
+      "learning_rate": 4.441621981359694e-05,
+      "loss": 0.7495,
+      "step": 6177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2385707488255893,
+      "learning_rate": 4.441359237219914e-05,
+      "loss": 0.6345,
+      "step": 6178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0640367039008463,
+      "learning_rate": 4.441096439052897e-05,
+      "loss": 1.5195,
+      "step": 6179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.832344574531742,
+      "learning_rate": 4.440833586865957e-05,
+      "loss": 1.1973,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.951767543861746,
+      "learning_rate": 4.440570680666408e-05,
+      "loss": 0.7404,
+      "step": 6181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5282129977345784,
+      "learning_rate": 4.4403077204615674e-05,
+      "loss": 0.6406,
+      "step": 6182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.025069521902276,
+      "learning_rate": 4.4400447062587527e-05,
+      "loss": 1.9829,
+      "step": 6183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.864490872827186,
+      "learning_rate": 4.4397816380652826e-05,
+      "loss": 1.041,
+      "step": 6184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.626888572434768,
+      "learning_rate": 4.43951851588848e-05,
+      "loss": 0.6588,
+      "step": 6185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 34.300656468640476,
+      "learning_rate": 4.4392553397356665e-05,
+      "loss": 0.9365,
+      "step": 6186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9343519183501314,
+      "learning_rate": 4.4389921096141666e-05,
+      "loss": 1.4512,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5972415449632904,
+      "learning_rate": 4.438728825531305e-05,
+      "loss": 1.1289,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.823336430457472,
+      "learning_rate": 4.438465487494411e-05,
+      "loss": 0.8564,
+      "step": 6189
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7801189051385884,
+      "learning_rate": 4.438202095510811e-05,
+      "loss": 0.8193,
+      "step": 6190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6556156971858966,
+      "learning_rate": 4.437938649587835e-05,
+      "loss": 0.9668,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.835796594529939,
+      "learning_rate": 4.437675149732815e-05,
+      "loss": 1.5098,
+      "step": 6192
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.820764225423615,
+      "learning_rate": 4.4374115959530856e-05,
+      "loss": 0.636,
+      "step": 6193
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9887372826074308,
+      "learning_rate": 4.4371479882559796e-05,
+      "loss": 0.8979,
+      "step": 6194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.276808167810201,
+      "learning_rate": 4.4368843266488334e-05,
+      "loss": 0.9336,
+      "step": 6195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3479010938667177,
+      "learning_rate": 4.436620611138985e-05,
+      "loss": 1.0098,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0365147440901157,
+      "learning_rate": 4.4363568417337743e-05,
+      "loss": 1.1758,
+      "step": 6197
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.640566714888521,
+      "learning_rate": 4.43609301844054e-05,
+      "loss": 0.8682,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.126553263881754,
+      "learning_rate": 4.435829141266625e-05,
+      "loss": 1.2393,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.219735971840209,
+      "learning_rate": 4.435565210219375e-05,
+      "loss": 0.9199,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8514864140641256,
+      "learning_rate": 4.435301225306131e-05,
+      "loss": 1.3711,
+      "step": 6201
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.963491339946185,
+      "learning_rate": 4.435037186534242e-05,
+      "loss": 1.5625,
+      "step": 6202
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.023835209458207,
+      "learning_rate": 4.434773093911056e-05,
+      "loss": 0.8615,
+      "step": 6203
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7350201548245763,
+      "learning_rate": 4.434508947443924e-05,
+      "loss": 0.692,
+      "step": 6204
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.1635336477459965,
+      "learning_rate": 4.4342447471401936e-05,
+      "loss": 0.9531,
+      "step": 6205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.02704895610592,
+      "learning_rate": 4.4339804930072205e-05,
+      "loss": 1.5332,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.585729400467146,
+      "learning_rate": 4.433716185052357e-05,
+      "loss": 1.4102,
+      "step": 6207
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4748797169457837,
+      "learning_rate": 4.4334518232829594e-05,
+      "loss": 0.8115,
+      "step": 6208
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9183153306662493,
+      "learning_rate": 4.433187407706385e-05,
+      "loss": 1.0146,
+      "step": 6209
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.63051131266164,
+      "learning_rate": 4.432922938329992e-05,
+      "loss": 1.1641,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.964361862806582,
+      "learning_rate": 4.43265841516114e-05,
+      "loss": 0.9292,
+      "step": 6211
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4285702605204618,
+      "learning_rate": 4.432393838207192e-05,
+      "loss": 1.2793,
+      "step": 6212
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.153211027732756,
+      "learning_rate": 4.4321292074755086e-05,
+      "loss": 0.9583,
+      "step": 6213
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.947408320042154,
+      "learning_rate": 4.4318645229734576e-05,
+      "loss": 0.9434,
+      "step": 6214
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4825066317065887,
+      "learning_rate": 4.4315997847084026e-05,
+      "loss": 1.0371,
+      "step": 6215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.802340317803015,
+      "learning_rate": 4.431334992687712e-05,
+      "loss": 1.1914,
+      "step": 6216
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.432898317174348,
+      "learning_rate": 4.431070146918754e-05,
+      "loss": 1.251,
+      "step": 6217
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.229033429194892,
+      "learning_rate": 4.4308052474089015e-05,
+      "loss": 1.3711,
+      "step": 6218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9296843493972573,
+      "learning_rate": 4.430540294165524e-05,
+      "loss": 1.0488,
+      "step": 6219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.528932240246234,
+      "learning_rate": 4.430275287195996e-05,
+      "loss": 0.9209,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7514177437822975,
+      "learning_rate": 4.430010226507693e-05,
+      "loss": 0.5732,
+      "step": 6221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8379630429399993,
+      "learning_rate": 4.429745112107991e-05,
+      "loss": 0.7432,
+      "step": 6222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1666829125506295,
+      "learning_rate": 4.4294799440042675e-05,
+      "loss": 0.8057,
+      "step": 6223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.084116415297962,
+      "learning_rate": 4.429214722203904e-05,
+      "loss": 0.9785,
+      "step": 6224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8353252349569975,
+      "learning_rate": 4.428949446714279e-05,
+      "loss": 0.948,
+      "step": 6225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.093599757763263,
+      "learning_rate": 4.428684117542777e-05,
+      "loss": 2.2695,
+      "step": 6226
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.751727996225767,
+      "learning_rate": 4.4284187346967805e-05,
+      "loss": 2.084,
+      "step": 6227
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.476186686613726,
+      "learning_rate": 4.428153298183676e-05,
+      "loss": 0.9775,
+      "step": 6228
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.744491782747554,
+      "learning_rate": 4.42788780801085e-05,
+      "loss": 0.9932,
+      "step": 6229
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1836617491779355,
+      "learning_rate": 4.427622264185691e-05,
+      "loss": 0.4956,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1266473624223874,
+      "learning_rate": 4.4273566667155895e-05,
+      "loss": 1.354,
+      "step": 6231
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0773620352168054,
+      "learning_rate": 4.427091015607937e-05,
+      "loss": 1.4883,
+      "step": 6232
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.334406665245428,
+      "learning_rate": 4.426825310870125e-05,
+      "loss": 1.5664,
+      "step": 6233
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.703978034573561,
+      "learning_rate": 4.4265595525095495e-05,
+      "loss": 1.2285,
+      "step": 6234
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.619898145185975,
+      "learning_rate": 4.426293740533606e-05,
+      "loss": 0.8408,
+      "step": 6235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.996314615986676,
+      "learning_rate": 4.426027874949692e-05,
+      "loss": 0.5591,
+      "step": 6236
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.234599765569415,
+      "learning_rate": 4.425761955765206e-05,
+      "loss": 0.6382,
+      "step": 6237
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1741309951643286,
+      "learning_rate": 4.425495982987549e-05,
+      "loss": 1.4766,
+      "step": 6238
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.210729151098479,
+      "learning_rate": 4.425229956624123e-05,
+      "loss": 0.8857,
+      "step": 6239
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.435884967699973,
+      "learning_rate": 4.4249638766823306e-05,
+      "loss": 0.9189,
+      "step": 6240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.719537603496661,
+      "learning_rate": 4.424697743169577e-05,
+      "loss": 1.082,
+      "step": 6241
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.418793343939339,
+      "learning_rate": 4.424431556093269e-05,
+      "loss": 1.0713,
+      "step": 6242
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8558644996990057,
+      "learning_rate": 4.424165315460814e-05,
+      "loss": 1.4766,
+      "step": 6243
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4141662970391087,
+      "learning_rate": 4.423899021279622e-05,
+      "loss": 0.5544,
+      "step": 6244
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8775279487734111,
+      "learning_rate": 4.4236326735571034e-05,
+      "loss": 0.9294,
+      "step": 6245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4654349217345497,
+      "learning_rate": 4.42336627230067e-05,
+      "loss": 1.3086,
+      "step": 6246
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8635736399164347,
+      "learning_rate": 4.4230998175177376e-05,
+      "loss": 0.9785,
+      "step": 6247
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4482237819349533,
+      "learning_rate": 4.422833309215719e-05,
+      "loss": 0.959,
+      "step": 6248
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6061983207292103,
+      "learning_rate": 4.422566747402033e-05,
+      "loss": 0.9443,
+      "step": 6249
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.811040662772767,
+      "learning_rate": 4.4223001320840965e-05,
+      "loss": 1.8066,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.963649986893191,
+      "learning_rate": 4.422033463269329e-05,
+      "loss": 0.918,
+      "step": 6251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6969180425256876,
+      "learning_rate": 4.421766740965154e-05,
+      "loss": 1.9238,
+      "step": 6252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.273355034077422,
+      "learning_rate": 4.421499965178992e-05,
+      "loss": 0.7949,
+      "step": 6253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.659761060581109,
+      "learning_rate": 4.42123313591827e-05,
+      "loss": 1.3223,
+      "step": 6254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.95398337954712,
+      "learning_rate": 4.4209662531904104e-05,
+      "loss": 1.2695,
+      "step": 6255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.658743581068869,
+      "learning_rate": 4.420699317002842e-05,
+      "loss": 0.9146,
+      "step": 6256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.763768415098771,
+      "learning_rate": 4.4204323273629944e-05,
+      "loss": 0.8193,
+      "step": 6257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.001200753554278,
+      "learning_rate": 4.420165284278296e-05,
+      "loss": 1.2773,
+      "step": 6258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.991894569552201,
+      "learning_rate": 4.419898187756179e-05,
+      "loss": 0.8838,
+      "step": 6259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.527130148245176,
+      "learning_rate": 4.419631037804079e-05,
+      "loss": 0.9893,
+      "step": 6260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4800643471091637,
+      "learning_rate": 4.4193638344294275e-05,
+      "loss": 0.8242,
+      "step": 6261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8971819767443105,
+      "learning_rate": 4.419096577639662e-05,
+      "loss": 0.7979,
+      "step": 6262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8163361560495703,
+      "learning_rate": 4.41882926744222e-05,
+      "loss": 1.6621,
+      "step": 6263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.725315877877154,
+      "learning_rate": 4.41856190384454e-05,
+      "loss": 0.79,
+      "step": 6264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.166148624005126,
+      "learning_rate": 4.4182944868540646e-05,
+      "loss": 1.5352,
+      "step": 6265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.559887390368541,
+      "learning_rate": 4.418027016478233e-05,
+      "loss": 1.2461,
+      "step": 6266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7243571721872204,
+      "learning_rate": 4.417759492724492e-05,
+      "loss": 1.1074,
+      "step": 6267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8868874421140662,
+      "learning_rate": 4.417491915600285e-05,
+      "loss": 1.1152,
+      "step": 6268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.052002489739275,
+      "learning_rate": 4.417224285113057e-05,
+      "loss": 1.5273,
+      "step": 6269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.094890485997707,
+      "learning_rate": 4.416956601270259e-05,
+      "loss": 1.4932,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.303899287906021,
+      "learning_rate": 4.416688864079338e-05,
+      "loss": 1.4131,
+      "step": 6271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1050574302540936,
+      "learning_rate": 4.4164210735477475e-05,
+      "loss": 1.0918,
+      "step": 6272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3658183453461237,
+      "learning_rate": 4.4161532296829376e-05,
+      "loss": 1.8281,
+      "step": 6273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.420634496725294,
+      "learning_rate": 4.415885332492364e-05,
+      "loss": 1.4102,
+      "step": 6274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9270018848233015,
+      "learning_rate": 4.415617381983481e-05,
+      "loss": 0.792,
+      "step": 6275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5809477620302768,
+      "learning_rate": 4.415349378163747e-05,
+      "loss": 0.8137,
+      "step": 6276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6657523336106057,
+      "learning_rate": 4.415081321040619e-05,
+      "loss": 1.1836,
+      "step": 6277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.632524931192541,
+      "learning_rate": 4.414813210621557e-05,
+      "loss": 1.0898,
+      "step": 6278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.683975053481682,
+      "learning_rate": 4.414545046914024e-05,
+      "loss": 1.5059,
+      "step": 6279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1582771392144617,
+      "learning_rate": 4.414276829925481e-05,
+      "loss": 0.8057,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.200573607532926,
+      "learning_rate": 4.414008559663393e-05,
+      "loss": 1.252,
+      "step": 6281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.064628347892595,
+      "learning_rate": 4.4137402361352266e-05,
+      "loss": 1.291,
+      "step": 6282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9714601694512504,
+      "learning_rate": 4.413471859348448e-05,
+      "loss": 1.4512,
+      "step": 6283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8763693160831907,
+      "learning_rate": 4.413203429310527e-05,
+      "loss": 1.2793,
+      "step": 6284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3925455165387404,
+      "learning_rate": 4.412934946028932e-05,
+      "loss": 1.1953,
+      "step": 6285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.54477853661086,
+      "learning_rate": 4.412666409511138e-05,
+      "loss": 1.2197,
+      "step": 6286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9882553997043666,
+      "learning_rate": 4.412397819764616e-05,
+      "loss": 1.4004,
+      "step": 6287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8958023124452565,
+      "learning_rate": 4.412129176796841e-05,
+      "loss": 0.9424,
+      "step": 6288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.005797922888694,
+      "learning_rate": 4.41186048061529e-05,
+      "loss": 0.5273,
+      "step": 6289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.29724003366446,
+      "learning_rate": 4.411591731227439e-05,
+      "loss": 1.4111,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.861294821154049,
+      "learning_rate": 4.411322928640769e-05,
+      "loss": 0.6714,
+      "step": 6291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8267593880356885,
+      "learning_rate": 4.41105407286276e-05,
+      "loss": 1.5,
+      "step": 6292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0283613352330274,
+      "learning_rate": 4.410785163900894e-05,
+      "loss": 1.0342,
+      "step": 6293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.801760827013869,
+      "learning_rate": 4.410516201762655e-05,
+      "loss": 1.1118,
+      "step": 6294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2141179909008044,
+      "learning_rate": 4.4102471864555275e-05,
+      "loss": 0.8467,
+      "step": 6295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.161202203366605,
+      "learning_rate": 4.409978117986999e-05,
+      "loss": 0.6504,
+      "step": 6296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.378047700068212,
+      "learning_rate": 4.4097089963645564e-05,
+      "loss": 1.0693,
+      "step": 6297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7444146396221476,
+      "learning_rate": 4.40943982159569e-05,
+      "loss": 0.9131,
+      "step": 6298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2815195608676375,
+      "learning_rate": 4.4091705936878915e-05,
+      "loss": 0.4805,
+      "step": 6299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.943525424822199,
+      "learning_rate": 4.4089013126486515e-05,
+      "loss": 0.8955,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.291120829704776,
+      "learning_rate": 4.4086319784854654e-05,
+      "loss": 1.6016,
+      "step": 6301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.670371720230452,
+      "learning_rate": 4.408362591205829e-05,
+      "loss": 1.1836,
+      "step": 6302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.199121092614436,
+      "learning_rate": 4.408093150817238e-05,
+      "loss": 1.3398,
+      "step": 6303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.129045668427024,
+      "learning_rate": 4.407823657327191e-05,
+      "loss": 0.8096,
+      "step": 6304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1736000949091605,
+      "learning_rate": 4.407554110743189e-05,
+      "loss": 0.9121,
+      "step": 6305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.412504478214574,
+      "learning_rate": 4.4072845110727324e-05,
+      "loss": 1.1328,
+      "step": 6306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1811379163741336,
+      "learning_rate": 4.4070148583233245e-05,
+      "loss": 0.7051,
+      "step": 6307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.303398590144675,
+      "learning_rate": 4.4067451525024696e-05,
+      "loss": 1.0137,
+      "step": 6308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9366658426457555,
+      "learning_rate": 4.406475393617673e-05,
+      "loss": 0.6157,
+      "step": 6309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.118235170609111,
+      "learning_rate": 4.406205581676442e-05,
+      "loss": 0.9185,
+      "step": 6310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.181821474443429,
+      "learning_rate": 4.405935716686286e-05,
+      "loss": 0.751,
+      "step": 6311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.656855794603439,
+      "learning_rate": 4.405665798654715e-05,
+      "loss": 0.9951,
+      "step": 6312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.485364425976274,
+      "learning_rate": 4.40539582758924e-05,
+      "loss": 0.9951,
+      "step": 6313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.769028116212261,
+      "learning_rate": 4.405125803497376e-05,
+      "loss": 0.9043,
+      "step": 6314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.44152606409961,
+      "learning_rate": 4.4048557263866355e-05,
+      "loss": 1.3691,
+      "step": 6315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.489348560901398,
+      "learning_rate": 4.4045855962645363e-05,
+      "loss": 1.498,
+      "step": 6316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.596104759447813,
+      "learning_rate": 4.404315413138595e-05,
+      "loss": 0.9902,
+      "step": 6317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3309328697026652,
+      "learning_rate": 4.404045177016331e-05,
+      "loss": 1.6895,
+      "step": 6318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5506746932902304,
+      "learning_rate": 4.403774887905265e-05,
+      "loss": 0.7866,
+      "step": 6319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6083381100286878,
+      "learning_rate": 4.4035045458129185e-05,
+      "loss": 0.96,
+      "step": 6320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.504367827876471,
+      "learning_rate": 4.403234150746815e-05,
+      "loss": 0.9785,
+      "step": 6321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.9283070029332725,
+      "learning_rate": 4.402963702714481e-05,
+      "loss": 1.6426,
+      "step": 6322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.933428522839432,
+      "learning_rate": 4.4026932017234414e-05,
+      "loss": 0.9121,
+      "step": 6323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.118853262126318,
+      "learning_rate": 4.402422647781225e-05,
+      "loss": 0.6104,
+      "step": 6324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7029327471438913,
+      "learning_rate": 4.40215204089536e-05,
+      "loss": 0.8154,
+      "step": 6325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1449502037970003,
+      "learning_rate": 4.4018813810733786e-05,
+      "loss": 1.2634,
+      "step": 6326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.530979000702268,
+      "learning_rate": 4.401610668322812e-05,
+      "loss": 0.8242,
+      "step": 6327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4152747474613183,
+      "learning_rate": 4.401339902651195e-05,
+      "loss": 1.2344,
+      "step": 6328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.170948168494495,
+      "learning_rate": 4.401069084066063e-05,
+      "loss": 1.1719,
+      "step": 6329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5798247851750666,
+      "learning_rate": 4.4007982125749515e-05,
+      "loss": 0.9131,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7069392993717827,
+      "learning_rate": 4.4005272881854e-05,
+      "loss": 1.4023,
+      "step": 6331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8026512638755245,
+      "learning_rate": 4.400256310904948e-05,
+      "loss": 1.2783,
+      "step": 6332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.907604637954315,
+      "learning_rate": 4.399985280741136e-05,
+      "loss": 1.043,
+      "step": 6333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6312593811229132,
+      "learning_rate": 4.399714197701508e-05,
+      "loss": 0.915,
+      "step": 6334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4621954715085055,
+      "learning_rate": 4.3994430617936063e-05,
+      "loss": 0.6919,
+      "step": 6335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.073006705683417,
+      "learning_rate": 4.3991718730249774e-05,
+      "loss": 1.3828,
+      "step": 6336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.089380859296176,
+      "learning_rate": 4.398900631403169e-05,
+      "loss": 1.0166,
+      "step": 6337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.489253903468853,
+      "learning_rate": 4.398629336935729e-05,
+      "loss": 1.3281,
+      "step": 6338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.783837903899796,
+      "learning_rate": 4.3983579896302075e-05,
+      "loss": 0.9551,
+      "step": 6339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.990028179890706,
+      "learning_rate": 4.398086589494156e-05,
+      "loss": 1.5312,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.168481482418112,
+      "learning_rate": 4.397815136535127e-05,
+      "loss": 0.835,
+      "step": 6341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.030238975946084,
+      "learning_rate": 4.3975436307606755e-05,
+      "loss": 1.1484,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0471634989091294,
+      "learning_rate": 4.397272072178357e-05,
+      "loss": 1.3809,
+      "step": 6343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.068389614100894,
+      "learning_rate": 4.397000460795729e-05,
+      "loss": 0.7332,
+      "step": 6344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.654280955587318,
+      "learning_rate": 4.3967287966203516e-05,
+      "loss": 0.9521,
+      "step": 6345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.8055605707650795,
+      "learning_rate": 4.396457079659783e-05,
+      "loss": 1.1289,
+      "step": 6346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.385418576790879,
+      "learning_rate": 4.396185309921586e-05,
+      "loss": 0.9561,
+      "step": 6347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6994923785533596,
+      "learning_rate": 4.395913487413324e-05,
+      "loss": 0.9609,
+      "step": 6348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.620997692911995,
+      "learning_rate": 4.395641612142562e-05,
+      "loss": 1.2832,
+      "step": 6349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6026393735065585,
+      "learning_rate": 4.395369684116865e-05,
+      "loss": 0.7588,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.690837028331638,
+      "learning_rate": 4.3950977033438005e-05,
+      "loss": 0.9912,
+      "step": 6351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.569933981832655,
+      "learning_rate": 4.39482566983094e-05,
+      "loss": 1.1406,
+      "step": 6352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.068840453599245,
+      "learning_rate": 4.3945535835858506e-05,
+      "loss": 0.9053,
+      "step": 6353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.57074861392049,
+      "learning_rate": 4.394281444616108e-05,
+      "loss": 1.3721,
+      "step": 6354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.80203836019698,
+      "learning_rate": 4.394009252929283e-05,
+      "loss": 0.7812,
+      "step": 6355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7226242873561013,
+      "learning_rate": 4.393737008532952e-05,
+      "loss": 0.7949,
+      "step": 6356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.208443154766647,
+      "learning_rate": 4.393464711434691e-05,
+      "loss": 1.3848,
+      "step": 6357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.291608932263906,
+      "learning_rate": 4.3931923616420775e-05,
+      "loss": 1.3848,
+      "step": 6358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3057281730815733,
+      "learning_rate": 4.3929199591626916e-05,
+      "loss": 1.125,
+      "step": 6359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9118545799415467,
+      "learning_rate": 4.3926475040041137e-05,
+      "loss": 1.584,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4555078630607694,
+      "learning_rate": 4.392374996173927e-05,
+      "loss": 0.5542,
+      "step": 6361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.875551088009639,
+      "learning_rate": 4.392102435679713e-05,
+      "loss": 1.5039,
+      "step": 6362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3311634078319976,
+      "learning_rate": 4.39182982252906e-05,
+      "loss": 0.959,
+      "step": 6363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7192499862969703,
+      "learning_rate": 4.391557156729553e-05,
+      "loss": 1.3457,
+      "step": 6364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8162898857006047,
+      "learning_rate": 4.39128443828878e-05,
+      "loss": 0.9492,
+      "step": 6365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.684028157218675,
+      "learning_rate": 4.3910116672143306e-05,
+      "loss": 0.969,
+      "step": 6366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.209431168281297,
+      "learning_rate": 4.390738843513797e-05,
+      "loss": 1.2334,
+      "step": 6367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.5851210558490605,
+      "learning_rate": 4.3904659671947715e-05,
+      "loss": 1.5156,
+      "step": 6368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1482485046488056,
+      "learning_rate": 4.390193038264846e-05,
+      "loss": 0.7583,
+      "step": 6369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.39646776316209,
+      "learning_rate": 4.389920056731619e-05,
+      "loss": 0.7593,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.916816745341132,
+      "learning_rate": 4.3896470226026866e-05,
+      "loss": 1.3936,
+      "step": 6371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2153530528569645,
+      "learning_rate": 4.389373935885646e-05,
+      "loss": 1.0371,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5806967190814354,
+      "learning_rate": 4.389100796588098e-05,
+      "loss": 0.7773,
+      "step": 6373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.622894220124335,
+      "learning_rate": 4.388827604717645e-05,
+      "loss": 0.7971,
+      "step": 6374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.452921211000689,
+      "learning_rate": 4.388554360281887e-05,
+      "loss": 1.0996,
+      "step": 6375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.897329113375099,
+      "learning_rate": 4.388281063288431e-05,
+      "loss": 0.7871,
+      "step": 6376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6310348263149597,
+      "learning_rate": 4.388007713744881e-05,
+      "loss": 1.25,
+      "step": 6377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.973765212962174,
+      "learning_rate": 4.387734311658845e-05,
+      "loss": 1.0898,
+      "step": 6378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4815019993648124,
+      "learning_rate": 4.3874608570379325e-05,
+      "loss": 0.8738,
+      "step": 6379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6893705024300427,
+      "learning_rate": 4.387187349889752e-05,
+      "loss": 0.8926,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.689839646907418,
+      "learning_rate": 4.386913790221917e-05,
+      "loss": 1.7207,
+      "step": 6381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2141348334125555,
+      "learning_rate": 4.3866401780420386e-05,
+      "loss": 0.8379,
+      "step": 6382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1195071965834815,
+      "learning_rate": 4.3863665133577315e-05,
+      "loss": 1.3398,
+      "step": 6383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.245888980985233,
+      "learning_rate": 4.386092796176613e-05,
+      "loss": 1.3213,
+      "step": 6384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.18407255913376,
+      "learning_rate": 4.3858190265062995e-05,
+      "loss": 0.7598,
+      "step": 6385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.477912526171209,
+      "learning_rate": 4.38554520435441e-05,
+      "loss": 1.2871,
+      "step": 6386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.753606932596437,
+      "learning_rate": 4.385271329728566e-05,
+      "loss": 1.5557,
+      "step": 6387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.383707188087763,
+      "learning_rate": 4.384997402636387e-05,
+      "loss": 0.8906,
+      "step": 6388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.496573006673133,
+      "learning_rate": 4.384723423085499e-05,
+      "loss": 0.7842,
+      "step": 6389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.221807205706263,
+      "learning_rate": 4.384449391083525e-05,
+      "loss": 1.0322,
+      "step": 6390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.0141496731271245,
+      "learning_rate": 4.384175306638091e-05,
+      "loss": 1.1172,
+      "step": 6391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8494916191602147,
+      "learning_rate": 4.383901169756825e-05,
+      "loss": 0.7148,
+      "step": 6392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.011975243976011,
+      "learning_rate": 4.383626980447358e-05,
+      "loss": 0.4448,
+      "step": 6393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.893183064220724,
+      "learning_rate": 4.383352738717318e-05,
+      "loss": 1.1025,
+      "step": 6394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.146461411619004,
+      "learning_rate": 4.383078444574338e-05,
+      "loss": 1.6836,
+      "step": 6395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0373503413555634,
+      "learning_rate": 4.382804098026051e-05,
+      "loss": 1.2168,
+      "step": 6396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.986672729968135,
+      "learning_rate": 4.382529699080094e-05,
+      "loss": 1.3516,
+      "step": 6397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1261347938288253,
+      "learning_rate": 4.3822552477441e-05,
+      "loss": 0.9497,
+      "step": 6398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.543255627179593,
+      "learning_rate": 4.38198074402571e-05,
+      "loss": 1.1016,
+      "step": 6399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.746886041289312,
+      "learning_rate": 4.3817061879325615e-05,
+      "loss": 0.8896,
+      "step": 6400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0962063064541523,
+      "learning_rate": 4.381431579472296e-05,
+      "loss": 0.6973,
+      "step": 6401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3798992420955254,
+      "learning_rate": 4.3811569186525546e-05,
+      "loss": 1.0889,
+      "step": 6402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.36033413543197,
+      "learning_rate": 4.3808822054809835e-05,
+      "loss": 1.7617,
+      "step": 6403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.666787940674778,
+      "learning_rate": 4.380607439965225e-05,
+      "loss": 0.8064,
+      "step": 6404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.410287120858087,
+      "learning_rate": 4.380332622112928e-05,
+      "loss": 0.5996,
+      "step": 6405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7793149753176856,
+      "learning_rate": 4.38005775193174e-05,
+      "loss": 0.7783,
+      "step": 6406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.568852013572318,
+      "learning_rate": 4.379782829429309e-05,
+      "loss": 0.7954,
+      "step": 6407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.072363045897049,
+      "learning_rate": 4.379507854613288e-05,
+      "loss": 0.6621,
+      "step": 6408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.3938878317675085,
+      "learning_rate": 4.379232827491328e-05,
+      "loss": 0.5852,
+      "step": 6409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.250866591366338,
+      "learning_rate": 4.378957748071083e-05,
+      "loss": 1.2988,
+      "step": 6410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.104565485056213,
+      "learning_rate": 4.37868261636021e-05,
+      "loss": 1.0195,
+      "step": 6411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8529103936019755,
+      "learning_rate": 4.378407432366364e-05,
+      "loss": 1.0381,
+      "step": 6412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5063867051281834,
+      "learning_rate": 4.3781321960972036e-05,
+      "loss": 0.417,
+      "step": 6413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.564162389656428,
+      "learning_rate": 4.3778569075603894e-05,
+      "loss": 0.7119,
+      "step": 6414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7529193814380073,
+      "learning_rate": 4.377581566763581e-05,
+      "loss": 1.0686,
+      "step": 6415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2387100527983717,
+      "learning_rate": 4.3773061737144425e-05,
+      "loss": 0.8594,
+      "step": 6416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.356393104477602,
+      "learning_rate": 4.377030728420638e-05,
+      "loss": 1.4424,
+      "step": 6417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.565844981047444,
+      "learning_rate": 4.376755230889832e-05,
+      "loss": 0.8862,
+      "step": 6418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.890077762835232,
+      "learning_rate": 4.376479681129691e-05,
+      "loss": 1.0918,
+      "step": 6419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3799310618486174,
+      "learning_rate": 4.3762040791478856e-05,
+      "loss": 1.5742,
+      "step": 6420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.826612964051014,
+      "learning_rate": 4.3759284249520845e-05,
+      "loss": 1.5117,
+      "step": 6421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.112131861474633,
+      "learning_rate": 4.3756527185499586e-05,
+      "loss": 1.1953,
+      "step": 6422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.674782314534262,
+      "learning_rate": 4.375376959949181e-05,
+      "loss": 1.1152,
+      "step": 6423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.149353594165402,
+      "learning_rate": 4.375101149157426e-05,
+      "loss": 1.1406,
+      "step": 6424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.594915748507417,
+      "learning_rate": 4.37482528618237e-05,
+      "loss": 1.2295,
+      "step": 6425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7921310112529536,
+      "learning_rate": 4.374549371031689e-05,
+      "loss": 1.1299,
+      "step": 6426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.3181740216243885,
+      "learning_rate": 4.374273403713063e-05,
+      "loss": 1.4902,
+      "step": 6427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.635588226966425,
+      "learning_rate": 4.37399738423417e-05,
+      "loss": 0.9609,
+      "step": 6428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.484871456889755,
+      "learning_rate": 4.373721312602693e-05,
+      "loss": 0.4893,
+      "step": 6429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.848878437758664,
+      "learning_rate": 4.373445188826315e-05,
+      "loss": 1.6191,
+      "step": 6430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8776776656808862,
+      "learning_rate": 4.37316901291272e-05,
+      "loss": 1.2695,
+      "step": 6431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5422791942411336,
+      "learning_rate": 4.372892784869593e-05,
+      "loss": 1.5098,
+      "step": 6432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.640440768963723,
+      "learning_rate": 4.3726165047046235e-05,
+      "loss": 1.6562,
+      "step": 6433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4239685853993507,
+      "learning_rate": 4.3723401724254984e-05,
+      "loss": 1.0332,
+      "step": 6434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4882605535344147,
+      "learning_rate": 4.37206378803991e-05,
+      "loss": 0.5234,
+      "step": 6435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.5471608362504385,
+      "learning_rate": 4.371787351555546e-05,
+      "loss": 0.874,
+      "step": 6436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.896842414812605,
+      "learning_rate": 4.371510862980104e-05,
+      "loss": 1.0146,
+      "step": 6437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.527854903119316,
+      "learning_rate": 4.3712343223212754e-05,
+      "loss": 1.0771,
+      "step": 6438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.493308511069059,
+      "learning_rate": 4.3709577295867575e-05,
+      "loss": 1.8203,
+      "step": 6439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.490315253329841,
+      "learning_rate": 4.370681084784248e-05,
+      "loss": 1.4458,
+      "step": 6440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.130196108856816,
+      "learning_rate": 4.3704043879214464e-05,
+      "loss": 1.0762,
+      "step": 6441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6155369393350179,
+      "learning_rate": 4.370127639006051e-05,
+      "loss": 0.3105,
+      "step": 6442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.206826299710452,
+      "learning_rate": 4.3698508380457646e-05,
+      "loss": 1.4248,
+      "step": 6443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.615508676948076,
+      "learning_rate": 4.369573985048291e-05,
+      "loss": 1.2344,
+      "step": 6444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2053676580704598,
+      "learning_rate": 4.369297080021334e-05,
+      "loss": 1.2656,
+      "step": 6445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.020252563814817,
+      "learning_rate": 4.3690201229726015e-05,
+      "loss": 0.9696,
+      "step": 6446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5855149214827993,
+      "learning_rate": 4.3687431139097985e-05,
+      "loss": 0.6494,
+      "step": 6447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.12958507592451,
+      "learning_rate": 4.368466052840636e-05,
+      "loss": 0.6011,
+      "step": 6448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0731872177563133,
+      "learning_rate": 4.368188939772824e-05,
+      "loss": 1.0596,
+      "step": 6449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9296064475639765,
+      "learning_rate": 4.367911774714074e-05,
+      "loss": 1.2148,
+      "step": 6450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.101928927084374,
+      "learning_rate": 4.3676345576721e-05,
+      "loss": 1.0723,
+      "step": 6451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0743684259644115,
+      "learning_rate": 4.3673572886546165e-05,
+      "loss": 1.6055,
+      "step": 6452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.261418007464941,
+      "learning_rate": 4.3670799676693394e-05,
+      "loss": 1.7109,
+      "step": 6453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.118483706143571,
+      "learning_rate": 4.366802594723988e-05,
+      "loss": 1.1562,
+      "step": 6454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7687856690817019,
+      "learning_rate": 4.3665251698262794e-05,
+      "loss": 0.7754,
+      "step": 6455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.393886454431711,
+      "learning_rate": 4.366247692983936e-05,
+      "loss": 0.8398,
+      "step": 6456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.147918860517307,
+      "learning_rate": 4.365970164204678e-05,
+      "loss": 1.1074,
+      "step": 6457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3257714991161693,
+      "learning_rate": 4.36569258349623e-05,
+      "loss": 0.5215,
+      "step": 6458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7709975134219933,
+      "learning_rate": 4.365414950866318e-05,
+      "loss": 1.0,
+      "step": 6459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8516055833987415,
+      "learning_rate": 4.3651372663226666e-05,
+      "loss": 1.0693,
+      "step": 6460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.98285641837171,
+      "learning_rate": 4.364859529873004e-05,
+      "loss": 0.584,
+      "step": 6461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6162277055966934,
+      "learning_rate": 4.364581741525061e-05,
+      "loss": 1.3906,
+      "step": 6462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.672663287253312,
+      "learning_rate": 4.364303901286566e-05,
+      "loss": 0.7263,
+      "step": 6463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0002268069680187,
+      "learning_rate": 4.3640260091652536e-05,
+      "loss": 0.9326,
+      "step": 6464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4300020756339795,
+      "learning_rate": 4.363748065168856e-05,
+      "loss": 1.0264,
+      "step": 6465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.441687958304287,
+      "learning_rate": 4.3634700693051076e-05,
+      "loss": 1.1445,
+      "step": 6466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3624534481245685,
+      "learning_rate": 4.363192021581746e-05,
+      "loss": 1.0742,
+      "step": 6467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5394148156501313,
+      "learning_rate": 4.3629139220065094e-05,
+      "loss": 0.6357,
+      "step": 6468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8772260167815875,
+      "learning_rate": 4.3626357705871365e-05,
+      "loss": 0.8828,
+      "step": 6469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9602106253975085,
+      "learning_rate": 4.3623575673313685e-05,
+      "loss": 1.8047,
+      "step": 6470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4443907596820806,
+      "learning_rate": 4.362079312246947e-05,
+      "loss": 1.3369,
+      "step": 6471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0295164519368685,
+      "learning_rate": 4.361801005341617e-05,
+      "loss": 0.4204,
+      "step": 6472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.102350840992646,
+      "learning_rate": 4.361522646623123e-05,
+      "loss": 1.4121,
+      "step": 6473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.031829792290223,
+      "learning_rate": 4.361244236099211e-05,
+      "loss": 1.4609,
+      "step": 6474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3415866277009862,
+      "learning_rate": 4.360965773777631e-05,
+      "loss": 0.7109,
+      "step": 6475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7560732276611355,
+      "learning_rate": 4.36068725966613e-05,
+      "loss": 0.8652,
+      "step": 6476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.147793456182699,
+      "learning_rate": 4.3604086937724595e-05,
+      "loss": 0.7998,
+      "step": 6477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.656610257337961,
+      "learning_rate": 4.3601300761043727e-05,
+      "loss": 0.8379,
+      "step": 6478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.496424632155669,
+      "learning_rate": 4.359851406669624e-05,
+      "loss": 1.2383,
+      "step": 6479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.652643335695991,
+      "learning_rate": 4.3595726854759674e-05,
+      "loss": 1.1309,
+      "step": 6480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.873948450298655,
+      "learning_rate": 4.359293912531159e-05,
+      "loss": 1.5442,
+      "step": 6481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.711814381231883,
+      "learning_rate": 4.359015087842959e-05,
+      "loss": 1.8027,
+      "step": 6482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.321761067651284,
+      "learning_rate": 4.358736211419126e-05,
+      "loss": 0.6914,
+      "step": 6483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1797324555385855,
+      "learning_rate": 4.358457283267421e-05,
+      "loss": 0.7463,
+      "step": 6484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.65970663847576,
+      "learning_rate": 4.358178303395606e-05,
+      "loss": 1.7715,
+      "step": 6485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.122847645105181,
+      "learning_rate": 4.3578992718114454e-05,
+      "loss": 0.7412,
+      "step": 6486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.533463374681838,
+      "learning_rate": 4.357620188522704e-05,
+      "loss": 1.2129,
+      "step": 6487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8667160825750613,
+      "learning_rate": 4.35734105353715e-05,
+      "loss": 0.625,
+      "step": 6488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.03645631703881,
+      "learning_rate": 4.357061866862549e-05,
+      "loss": 0.7549,
+      "step": 6489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3838943266906227,
+      "learning_rate": 4.356782628506674e-05,
+      "loss": 1.1748,
+      "step": 6490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8309643738715575,
+      "learning_rate": 4.3565033384772926e-05,
+      "loss": 1.3379,
+      "step": 6491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.615646092701525,
+      "learning_rate": 4.3562239967821805e-05,
+      "loss": 1.332,
+      "step": 6492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.322116898724737,
+      "learning_rate": 4.355944603429109e-05,
+      "loss": 0.9663,
+      "step": 6493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.451459088159374,
+      "learning_rate": 4.355665158425856e-05,
+      "loss": 1.1172,
+      "step": 6494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5914147792446847,
+      "learning_rate": 4.355385661780197e-05,
+      "loss": 1.167,
+      "step": 6495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.391280743983208,
+      "learning_rate": 4.355106113499909e-05,
+      "loss": 0.9956,
+      "step": 6496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9031995946459723,
+      "learning_rate": 4.354826513592774e-05,
+      "loss": 1.0488,
+      "step": 6497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9053840577679786,
+      "learning_rate": 4.354546862066573e-05,
+      "loss": 1.0654,
+      "step": 6498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3793284598931415,
+      "learning_rate": 4.354267158929087e-05,
+      "loss": 0.7939,
+      "step": 6499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5020205251374366,
+      "learning_rate": 4.3539874041881004e-05,
+      "loss": 0.894,
+      "step": 6500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7573686193880635,
+      "learning_rate": 4.3537075978513994e-05,
+      "loss": 0.6514,
+      "step": 6501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.238182467526824,
+      "learning_rate": 4.353427739926771e-05,
+      "loss": 0.6021,
+      "step": 6502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.6837826673694964,
+      "learning_rate": 4.353147830422003e-05,
+      "loss": 1.1973,
+      "step": 6503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.705834238302653,
+      "learning_rate": 4.3528678693448854e-05,
+      "loss": 0.8423,
+      "step": 6504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6550501670033606,
+      "learning_rate": 4.352587856703209e-05,
+      "loss": 0.6924,
+      "step": 6505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6779168064736063,
+      "learning_rate": 4.352307792504767e-05,
+      "loss": 1.251,
+      "step": 6506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.759341747615693,
+      "learning_rate": 4.352027676757353e-05,
+      "loss": 0.9502,
+      "step": 6507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.46114317019411,
+      "learning_rate": 4.351747509468763e-05,
+      "loss": 1.2246,
+      "step": 6508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8991815069669844,
+      "learning_rate": 4.351467290646794e-05,
+      "loss": 1.5762,
+      "step": 6509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.3007788108105025,
+      "learning_rate": 4.351187020299242e-05,
+      "loss": 1.0576,
+      "step": 6510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.437094073784078,
+      "learning_rate": 4.350906698433911e-05,
+      "loss": 1.2754,
+      "step": 6511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8546336736338924,
+      "learning_rate": 4.3506263250585996e-05,
+      "loss": 0.1812,
+      "step": 6512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.488041990059921,
+      "learning_rate": 4.35034590018111e-05,
+      "loss": 1.3652,
+      "step": 6513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5621711706056858,
+      "learning_rate": 4.350065423809249e-05,
+      "loss": 1.3906,
+      "step": 6514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.495641929268964,
+      "learning_rate": 4.349784895950819e-05,
+      "loss": 1.7168,
+      "step": 6515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.943114828192703,
+      "learning_rate": 4.3495043166136284e-05,
+      "loss": 1.2109,
+      "step": 6516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7150410751973064,
+      "learning_rate": 4.349223685805486e-05,
+      "loss": 1.4268,
+      "step": 6517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.761167220277497,
+      "learning_rate": 4.3489430035342015e-05,
+      "loss": 1.2598,
+      "step": 6518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.70335626158307,
+      "learning_rate": 4.348662269807586e-05,
+      "loss": 0.7921,
+      "step": 6519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.969617536174857,
+      "learning_rate": 4.3483814846334515e-05,
+      "loss": 1.3438,
+      "step": 6520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.146475933360334,
+      "learning_rate": 4.3481006480196126e-05,
+      "loss": 1.3945,
+      "step": 6521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.372953738044973,
+      "learning_rate": 4.347819759973886e-05,
+      "loss": 1.3398,
+      "step": 6522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.460395971389372,
+      "learning_rate": 4.3475388205040874e-05,
+      "loss": 1.4375,
+      "step": 6523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.8825822114035455,
+      "learning_rate": 4.347257829618036e-05,
+      "loss": 1.4688,
+      "step": 6524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.446848089548209,
+      "learning_rate": 4.34697678732355e-05,
+      "loss": 1.3867,
+      "step": 6525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.970554812504671,
+      "learning_rate": 4.346695693628452e-05,
+      "loss": 1.2329,
+      "step": 6526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7156319248401224,
+      "learning_rate": 4.3464145485405653e-05,
+      "loss": 0.731,
+      "step": 6527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2295050200846194,
+      "learning_rate": 4.3461333520677136e-05,
+      "loss": 1.4072,
+      "step": 6528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.491366036741825,
+      "learning_rate": 4.345852104217722e-05,
+      "loss": 1.8984,
+      "step": 6529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.18117344270142,
+      "learning_rate": 4.345570804998418e-05,
+      "loss": 0.9072,
+      "step": 6530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.022053012941549,
+      "learning_rate": 4.345289454417629e-05,
+      "loss": 1.1309,
+      "step": 6531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.623707870349563,
+      "learning_rate": 4.3450080524831864e-05,
+      "loss": 1.9688,
+      "step": 6532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2386106877775522,
+      "learning_rate": 4.344726599202921e-05,
+      "loss": 1.2754,
+      "step": 6533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9896594490879718,
+      "learning_rate": 4.344445094584665e-05,
+      "loss": 1.3457,
+      "step": 6534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.833498878623554,
+      "learning_rate": 4.344163538636252e-05,
+      "loss": 1.1699,
+      "step": 6535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.597880282447653,
+      "learning_rate": 4.3438819313655196e-05,
+      "loss": 1.2236,
+      "step": 6536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.247523315437544,
+      "learning_rate": 4.343600272780304e-05,
+      "loss": 1.2666,
+      "step": 6537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.171926374308079,
+      "learning_rate": 4.343318562888442e-05,
+      "loss": 0.8867,
+      "step": 6538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.920475836593051,
+      "learning_rate": 4.343036801697776e-05,
+      "loss": 1.9629,
+      "step": 6539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.095064141559486,
+      "learning_rate": 4.342754989216146e-05,
+      "loss": 0.959,
+      "step": 6540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.365751932787758,
+      "learning_rate": 4.342473125451394e-05,
+      "loss": 1.2441,
+      "step": 6541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2475189128557003,
+      "learning_rate": 4.3421912104113656e-05,
+      "loss": 1.3184,
+      "step": 6542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.314579073097888,
+      "learning_rate": 4.3419092441039055e-05,
+      "loss": 0.7842,
+      "step": 6543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4269679510061897,
+      "learning_rate": 4.3416272265368615e-05,
+      "loss": 1.5312,
+      "step": 6544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.43997330994939,
+      "learning_rate": 4.3413451577180806e-05,
+      "loss": 0.9663,
+      "step": 6545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.579741580607716,
+      "learning_rate": 4.3410630376554144e-05,
+      "loss": 1.6797,
+      "step": 6546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7234251668785396,
+      "learning_rate": 4.340780866356713e-05,
+      "loss": 0.8755,
+      "step": 6547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.187336615864109,
+      "learning_rate": 4.34049864382983e-05,
+      "loss": 0.5537,
+      "step": 6548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.389626691626686,
+      "learning_rate": 4.340216370082618e-05,
+      "loss": 1.7441,
+      "step": 6549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8712008417833905,
+      "learning_rate": 4.339934045122935e-05,
+      "loss": 0.9102,
+      "step": 6550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9721472325674085,
+      "learning_rate": 4.3396516689586355e-05,
+      "loss": 0.8086,
+      "step": 6551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.6909014758235745,
+      "learning_rate": 4.3393692415975796e-05,
+      "loss": 0.6279,
+      "step": 6552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.435665127359178,
+      "learning_rate": 4.339086763047626e-05,
+      "loss": 1.5957,
+      "step": 6553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7286761379336975,
+      "learning_rate": 4.338804233316636e-05,
+      "loss": 1.7695,
+      "step": 6554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.862907867192703,
+      "learning_rate": 4.338521652412474e-05,
+      "loss": 1.25,
+      "step": 6555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3916526031966767,
+      "learning_rate": 4.338239020343002e-05,
+      "loss": 0.6113,
+      "step": 6556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.473435358351452,
+      "learning_rate": 4.337956337116087e-05,
+      "loss": 1.2949,
+      "step": 6557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.455112405445087,
+      "learning_rate": 4.3376736027395945e-05,
+      "loss": 1.4824,
+      "step": 6558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6402275949106566,
+      "learning_rate": 4.337390817221394e-05,
+      "loss": 1.0352,
+      "step": 6559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0110377074800474,
+      "learning_rate": 4.3371079805693557e-05,
+      "loss": 0.8877,
+      "step": 6560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.321706436630839,
+      "learning_rate": 4.33682509279135e-05,
+      "loss": 1.209,
+      "step": 6561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.469075664719622,
+      "learning_rate": 4.3365421538952496e-05,
+      "loss": 1.6963,
+      "step": 6562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.539372812717294,
+      "learning_rate": 4.3362591638889284e-05,
+      "loss": 1.1406,
+      "step": 6563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.299713196172035,
+      "learning_rate": 4.3359761227802624e-05,
+      "loss": 0.7188,
+      "step": 6564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1274914441233634,
+      "learning_rate": 4.3356930305771284e-05,
+      "loss": 0.8442,
+      "step": 6565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.407440890361023,
+      "learning_rate": 4.335409887287405e-05,
+      "loss": 1.0293,
+      "step": 6566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9920147989251895,
+      "learning_rate": 4.3351266929189715e-05,
+      "loss": 0.6371,
+      "step": 6567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6210117924209198,
+      "learning_rate": 4.334843447479708e-05,
+      "loss": 1.0054,
+      "step": 6568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.796824364230444,
+      "learning_rate": 4.3345601509775005e-05,
+      "loss": 0.9668,
+      "step": 6569
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.541511906374568,
+      "learning_rate": 4.334276803420229e-05,
+      "loss": 1.4609,
+      "step": 6570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3102965295880233,
+      "learning_rate": 4.3339934048157816e-05,
+      "loss": 1.2168,
+      "step": 6571
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5339934940329885,
+      "learning_rate": 4.333709955172044e-05,
+      "loss": 0.9385,
+      "step": 6572
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7658099096520257,
+      "learning_rate": 4.3334264544969056e-05,
+      "loss": 1.3662,
+      "step": 6573
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.837164949549013,
+      "learning_rate": 4.333142902798255e-05,
+      "loss": 1.9062,
+      "step": 6574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.253952790389877,
+      "learning_rate": 4.3328593000839835e-05,
+      "loss": 0.8203,
+      "step": 6575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.939059391456108,
+      "learning_rate": 4.3325756463619835e-05,
+      "loss": 0.8232,
+      "step": 6576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.4950129107161505,
+      "learning_rate": 4.3322919416401495e-05,
+      "loss": 1.0591,
+      "step": 6577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8705067520162495,
+      "learning_rate": 4.332008185926377e-05,
+      "loss": 0.8076,
+      "step": 6578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.377928840524612,
+      "learning_rate": 4.331724379228563e-05,
+      "loss": 0.8682,
+      "step": 6579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.380530431322354,
+      "learning_rate": 4.331440521554604e-05,
+      "loss": 1.3945,
+      "step": 6580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0975147806697043,
+      "learning_rate": 4.3311566129124014e-05,
+      "loss": 0.9941,
+      "step": 6581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.659321073921486,
+      "learning_rate": 4.330872653309857e-05,
+      "loss": 0.708,
+      "step": 6582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.08453228983285,
+      "learning_rate": 4.3305886427548706e-05,
+      "loss": 1.0217,
+      "step": 6583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8111502905600267,
+      "learning_rate": 4.3303045812553467e-05,
+      "loss": 1.5039,
+      "step": 6584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6190072636570947,
+      "learning_rate": 4.330020468819193e-05,
+      "loss": 0.3491,
+      "step": 6585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6269607487110567,
+      "learning_rate": 4.329736305454314e-05,
+      "loss": 1.3818,
+      "step": 6586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.598400377964422,
+      "learning_rate": 4.3294520911686184e-05,
+      "loss": 0.8291,
+      "step": 6587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.579904880785301,
+      "learning_rate": 4.3291678259700163e-05,
+      "loss": 1.4004,
+      "step": 6588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.715571595469066,
+      "learning_rate": 4.3288835098664173e-05,
+      "loss": 1.3789,
+      "step": 6589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0479520337403465,
+      "learning_rate": 4.3285991428657354e-05,
+      "loss": 0.7185,
+      "step": 6590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8194990596482166,
+      "learning_rate": 4.3283147249758846e-05,
+      "loss": 0.5659,
+      "step": 6591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6398051416848314,
+      "learning_rate": 4.328030256204778e-05,
+      "loss": 1.1836,
+      "step": 6592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.588839235157737,
+      "learning_rate": 4.327745736560335e-05,
+      "loss": 1.3284,
+      "step": 6593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.632928005556219,
+      "learning_rate": 4.32746116605047e-05,
+      "loss": 0.9395,
+      "step": 6594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.350004235627315,
+      "learning_rate": 4.327176544683107e-05,
+      "loss": 1.1348,
+      "step": 6595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1688933179480943,
+      "learning_rate": 4.3268918724661634e-05,
+      "loss": 0.3142,
+      "step": 6596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.11406057832036,
+      "learning_rate": 4.3266071494075636e-05,
+      "loss": 1.5605,
+      "step": 6597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2388584188602434,
+      "learning_rate": 4.3263223755152294e-05,
+      "loss": 0.7656,
+      "step": 6598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8340183281816116,
+      "learning_rate": 4.326037550797087e-05,
+      "loss": 0.7554,
+      "step": 6599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.610262331005994,
+      "learning_rate": 4.3257526752610635e-05,
+      "loss": 1.4805,
+      "step": 6600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.755067404668415,
+      "learning_rate": 4.325467748915086e-05,
+      "loss": 1.2539,
+      "step": 6601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.756638809009756,
+      "learning_rate": 4.325182771767084e-05,
+      "loss": 0.1833,
+      "step": 6602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.556439000536558,
+      "learning_rate": 4.3248977438249896e-05,
+      "loss": 0.9727,
+      "step": 6603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.879012829563907,
+      "learning_rate": 4.3246126650967334e-05,
+      "loss": 1.4482,
+      "step": 6604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5482113375162294,
+      "learning_rate": 4.3243275355902494e-05,
+      "loss": 0.6455,
+      "step": 6605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9857802353131127,
+      "learning_rate": 4.3240423553134724e-05,
+      "loss": 1.0391,
+      "step": 6606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.355636315816267,
+      "learning_rate": 4.323757124274339e-05,
+      "loss": 1.3047,
+      "step": 6607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.269563815856608,
+      "learning_rate": 4.3234718424807884e-05,
+      "loss": 0.9883,
+      "step": 6608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.243889860397791,
+      "learning_rate": 4.3231865099407576e-05,
+      "loss": 1.1777,
+      "step": 6609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.85717355166705,
+      "learning_rate": 4.3229011266621895e-05,
+      "loss": 0.9932,
+      "step": 6610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8114955391865877,
+      "learning_rate": 4.3226156926530245e-05,
+      "loss": 1.3877,
+      "step": 6611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.8843027686792535,
+      "learning_rate": 4.322330207921207e-05,
+      "loss": 0.8613,
+      "step": 6612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.171662244736138,
+      "learning_rate": 4.322044672474681e-05,
+      "loss": 0.9561,
+      "step": 6613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.067631393418795,
+      "learning_rate": 4.321759086321394e-05,
+      "loss": 1.208,
+      "step": 6614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7481744081880657,
+      "learning_rate": 4.3214734494692935e-05,
+      "loss": 0.6992,
+      "step": 6615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.788409451230236,
+      "learning_rate": 4.321187761926328e-05,
+      "loss": 1.0176,
+      "step": 6616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.661464554293554,
+      "learning_rate": 4.3209020237004495e-05,
+      "loss": 0.9639,
+      "step": 6617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.523353378225248,
+      "learning_rate": 4.320616234799608e-05,
+      "loss": 1.3301,
+      "step": 6618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.971156571907034,
+      "learning_rate": 4.3203303952317586e-05,
+      "loss": 1.2871,
+      "step": 6619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2904926410216735,
+      "learning_rate": 4.3200445050048545e-05,
+      "loss": 1.2695,
+      "step": 6620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.285465473869815,
+      "learning_rate": 4.319758564126853e-05,
+      "loss": 1.3652,
+      "step": 6621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.251928244963655,
+      "learning_rate": 4.319472572605712e-05,
+      "loss": 1.0752,
+      "step": 6622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.217876017893168,
+      "learning_rate": 4.31918653044939e-05,
+      "loss": 1.377,
+      "step": 6623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.001664065450121,
+      "learning_rate": 4.318900437665846e-05,
+      "loss": 1.2793,
+      "step": 6624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8877068713836973,
+      "learning_rate": 4.3186142942630446e-05,
+      "loss": 0.9795,
+      "step": 6625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.918231015921396,
+      "learning_rate": 4.3183281002489476e-05,
+      "loss": 1.0664,
+      "step": 6626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7970208034071065,
+      "learning_rate": 4.3180418556315194e-05,
+      "loss": 0.8438,
+      "step": 6627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.051538321249682,
+      "learning_rate": 4.317755560418727e-05,
+      "loss": 0.7671,
+      "step": 6628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.616059399452357,
+      "learning_rate": 4.317469214618537e-05,
+      "loss": 1.3477,
+      "step": 6629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4200743841924393,
+      "learning_rate": 4.317182818238919e-05,
+      "loss": 1.3027,
+      "step": 6630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.020493484497278,
+      "learning_rate": 4.316896371287843e-05,
+      "loss": 0.9092,
+      "step": 6631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.71297089130263,
+      "learning_rate": 4.31660987377328e-05,
+      "loss": 0.9951,
+      "step": 6632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.902663748582503,
+      "learning_rate": 4.3163233257032043e-05,
+      "loss": 1.2969,
+      "step": 6633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.097967169038243,
+      "learning_rate": 4.316036727085589e-05,
+      "loss": 1.791,
+      "step": 6634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.495665512004531,
+      "learning_rate": 4.315750077928413e-05,
+      "loss": 0.8257,
+      "step": 6635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.553772257259844,
+      "learning_rate": 4.31546337823965e-05,
+      "loss": 0.9551,
+      "step": 6636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.826107054198577,
+      "learning_rate": 4.31517662802728e-05,
+      "loss": 1.9395,
+      "step": 6637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1683539520894066,
+      "learning_rate": 4.314889827299284e-05,
+      "loss": 0.7251,
+      "step": 6638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7960043328826263,
+      "learning_rate": 4.314602976063642e-05,
+      "loss": 1.0977,
+      "step": 6639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.784012759802966,
+      "learning_rate": 4.314316074328339e-05,
+      "loss": 1.6777,
+      "step": 6640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.378123488156293,
+      "learning_rate": 4.314029122101358e-05,
+      "loss": 1.1699,
+      "step": 6641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.059329980785899,
+      "learning_rate": 4.313742119390685e-05,
+      "loss": 1.3574,
+      "step": 6642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.662071531515344,
+      "learning_rate": 4.3134550662043074e-05,
+      "loss": 0.8955,
+      "step": 6643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1540540487438258,
+      "learning_rate": 4.3131679625502134e-05,
+      "loss": 0.6987,
+      "step": 6644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1842725650634285,
+      "learning_rate": 4.312880808436393e-05,
+      "loss": 0.7998,
+      "step": 6645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.562552634000893,
+      "learning_rate": 4.3125936038708374e-05,
+      "loss": 1.1909,
+      "step": 6646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7760270185722287,
+      "learning_rate": 4.31230634886154e-05,
+      "loss": 0.8276,
+      "step": 6647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1618850715713602,
+      "learning_rate": 4.312019043416495e-05,
+      "loss": 0.7744,
+      "step": 6648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.887122603861521,
+      "learning_rate": 4.3117316875436976e-05,
+      "loss": 0.6377,
+      "step": 6649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.018849949680196,
+      "learning_rate": 4.3114442812511445e-05,
+      "loss": 0.75,
+      "step": 6650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1037294921963396,
+      "learning_rate": 4.311156824546835e-05,
+      "loss": 0.9688,
+      "step": 6651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5983396620690686,
+      "learning_rate": 4.310869317438768e-05,
+      "loss": 0.6978,
+      "step": 6652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1545792893422373,
+      "learning_rate": 4.310581759934945e-05,
+      "loss": 0.4146,
+      "step": 6653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2031784425324465,
+      "learning_rate": 4.3102941520433695e-05,
+      "loss": 0.8652,
+      "step": 6654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.01888893563601,
+      "learning_rate": 4.310006493772044e-05,
+      "loss": 0.9087,
+      "step": 6655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.727801039146722,
+      "learning_rate": 4.3097187851289744e-05,
+      "loss": 0.6489,
+      "step": 6656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.618507257468608,
+      "learning_rate": 4.309431026122169e-05,
+      "loss": 0.7041,
+      "step": 6657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.528527762226103,
+      "learning_rate": 4.309143216759634e-05,
+      "loss": 0.6597,
+      "step": 6658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.154341551150068,
+      "learning_rate": 4.30885535704938e-05,
+      "loss": 1.4531,
+      "step": 6659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.381694793978482,
+      "learning_rate": 4.308567446999416e-05,
+      "loss": 0.9692,
+      "step": 6660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7174483943415417,
+      "learning_rate": 4.308279486617758e-05,
+      "loss": 1.5664,
+      "step": 6661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.414943003643346,
+      "learning_rate": 4.307991475912419e-05,
+      "loss": 0.8552,
+      "step": 6662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.313753561535469,
+      "learning_rate": 4.3077034148914105e-05,
+      "loss": 0.9453,
+      "step": 6663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8873293673478466,
+      "learning_rate": 4.307415303562754e-05,
+      "loss": 1.0078,
+      "step": 6664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.614069408339228,
+      "learning_rate": 4.307127141934465e-05,
+      "loss": 1.1953,
+      "step": 6665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.325718840909634,
+      "learning_rate": 4.306838930014563e-05,
+      "loss": 1.4648,
+      "step": 6666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2788268506038274,
+      "learning_rate": 4.306550667811069e-05,
+      "loss": 1.25,
+      "step": 6667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.971460329923583,
+      "learning_rate": 4.306262355332006e-05,
+      "loss": 1.0732,
+      "step": 6668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1792398273520948,
+      "learning_rate": 4.305973992585397e-05,
+      "loss": 0.8159,
+      "step": 6669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8123197794796762,
+      "learning_rate": 4.305685579579266e-05,
+      "loss": 1.084,
+      "step": 6670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6114773989390905,
+      "learning_rate": 4.305397116321641e-05,
+      "loss": 1.2969,
+      "step": 6671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4264680876382063,
+      "learning_rate": 4.305108602820549e-05,
+      "loss": 1.3711,
+      "step": 6672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.923751962570569,
+      "learning_rate": 4.30482003908402e-05,
+      "loss": 1.2012,
+      "step": 6673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.96085034035637,
+      "learning_rate": 4.304531425120084e-05,
+      "loss": 1.3379,
+      "step": 6674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.77215748808277,
+      "learning_rate": 4.3042427609367734e-05,
+      "loss": 0.7936,
+      "step": 6675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.015103436155456,
+      "learning_rate": 4.30395404654212e-05,
+      "loss": 0.8691,
+      "step": 6676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.306498234961846,
+      "learning_rate": 4.3036652819441604e-05,
+      "loss": 1.0781,
+      "step": 6677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.101683405956061,
+      "learning_rate": 4.3033764671509305e-05,
+      "loss": 1.4766,
+      "step": 6678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.230361602722878,
+      "learning_rate": 4.303087602170467e-05,
+      "loss": 0.8701,
+      "step": 6679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7753044864559073,
+      "learning_rate": 4.302798687010811e-05,
+      "loss": 0.9648,
+      "step": 6680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.7321074365422096,
+      "learning_rate": 4.30250972168e-05,
+      "loss": 0.959,
+      "step": 6681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.455610825946045,
+      "learning_rate": 4.302220706186078e-05,
+      "loss": 1.332,
+      "step": 6682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.650184077941126,
+      "learning_rate": 4.3019316405370873e-05,
+      "loss": 0.9033,
+      "step": 6683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0184330491506004,
+      "learning_rate": 4.301642524741073e-05,
+      "loss": 1.1035,
+      "step": 6684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.976939390179305,
+      "learning_rate": 4.3013533588060804e-05,
+      "loss": 1.3359,
+      "step": 6685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3728471334404917,
+      "learning_rate": 4.3010641427401566e-05,
+      "loss": 0.835,
+      "step": 6686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.132206200219101,
+      "learning_rate": 4.300774876551352e-05,
+      "loss": 1.6504,
+      "step": 6687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.032494243044339,
+      "learning_rate": 4.300485560247714e-05,
+      "loss": 1.5977,
+      "step": 6688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3526708075916707,
+      "learning_rate": 4.300196193837298e-05,
+      "loss": 0.9482,
+      "step": 6689
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.955348443151392,
+      "learning_rate": 4.2999067773281534e-05,
+      "loss": 0.6152,
+      "step": 6690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.950688029376218,
+      "learning_rate": 4.299617310728336e-05,
+      "loss": 0.9814,
+      "step": 6691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.555360384643241,
+      "learning_rate": 4.299327794045902e-05,
+      "loss": 0.4419,
+      "step": 6692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.939988873117866,
+      "learning_rate": 4.299038227288907e-05,
+      "loss": 1.0596,
+      "step": 6693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0141643880705957,
+      "learning_rate": 4.2987486104654115e-05,
+      "loss": 1.2261,
+      "step": 6694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8909507523172286,
+      "learning_rate": 4.298458943583474e-05,
+      "loss": 0.5168,
+      "step": 6695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.430365006452662,
+      "learning_rate": 4.2981692266511565e-05,
+      "loss": 1.9277,
+      "step": 6696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.666296703752896,
+      "learning_rate": 4.297879459676522e-05,
+      "loss": 1.3555,
+      "step": 6697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.70589750815307,
+      "learning_rate": 4.297589642667633e-05,
+      "loss": 1.416,
+      "step": 6698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.145403023373757,
+      "learning_rate": 4.297299775632556e-05,
+      "loss": 1.2998,
+      "step": 6699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5052789267086313,
+      "learning_rate": 4.297009858579359e-05,
+      "loss": 1.1035,
+      "step": 6700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.538848022792407,
+      "learning_rate": 4.296719891516108e-05,
+      "loss": 1.2197,
+      "step": 6701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.70889140901746,
+      "learning_rate": 4.296429874450875e-05,
+      "loss": 1.0603,
+      "step": 6702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.945964253290269,
+      "learning_rate": 4.296139807391729e-05,
+      "loss": 0.874,
+      "step": 6703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.193297969828704,
+      "learning_rate": 4.295849690346744e-05,
+      "loss": 0.9058,
+      "step": 6704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.7158702833167165,
+      "learning_rate": 4.295559523323993e-05,
+      "loss": 0.9424,
+      "step": 6705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8359474202973014,
+      "learning_rate": 4.295269306331552e-05,
+      "loss": 1.2852,
+      "step": 6706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.084014215983319,
+      "learning_rate": 4.294979039377496e-05,
+      "loss": 1.4414,
+      "step": 6707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.898019094408916,
+      "learning_rate": 4.294688722469904e-05,
+      "loss": 0.9756,
+      "step": 6708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.288407380475004,
+      "learning_rate": 4.294398355616856e-05,
+      "loss": 1.042,
+      "step": 6709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2262208910778782,
+      "learning_rate": 4.294107938826433e-05,
+      "loss": 0.8545,
+      "step": 6710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3314274656436074,
+      "learning_rate": 4.293817472106715e-05,
+      "loss": 0.6719,
+      "step": 6711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3657861974664582,
+      "learning_rate": 4.293526955465787e-05,
+      "loss": 1.3975,
+      "step": 6712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.248100293235142,
+      "learning_rate": 4.2932363889117354e-05,
+      "loss": 1.6797,
+      "step": 6713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1034676876411043,
+      "learning_rate": 4.292945772452644e-05,
+      "loss": 1.0,
+      "step": 6714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.682152214635155,
+      "learning_rate": 4.2926551060966026e-05,
+      "loss": 0.9731,
+      "step": 6715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8130402787813735,
+      "learning_rate": 4.2923643898516986e-05,
+      "loss": 0.7255,
+      "step": 6716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.1796813898403,
+      "learning_rate": 4.292073623726024e-05,
+      "loss": 0.7666,
+      "step": 6717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.350693895090117,
+      "learning_rate": 4.291782807727669e-05,
+      "loss": 0.7822,
+      "step": 6718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.163740249388749,
+      "learning_rate": 4.2914919418647284e-05,
+      "loss": 0.9102,
+      "step": 6719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3830776379768848,
+      "learning_rate": 4.2912010261452975e-05,
+      "loss": 1.4727,
+      "step": 6720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.215092182410023,
+      "learning_rate": 4.29091006057747e-05,
+      "loss": 1.9023,
+      "step": 6721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0426689656466537,
+      "learning_rate": 4.2906190451693437e-05,
+      "loss": 0.9966,
+      "step": 6722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8409926923185127,
+      "learning_rate": 4.2903279799290195e-05,
+      "loss": 1.0059,
+      "step": 6723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.389060396687597,
+      "learning_rate": 4.290036864864597e-05,
+      "loss": 1.4746,
+      "step": 6724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.451644936920621,
+      "learning_rate": 4.289745699984177e-05,
+      "loss": 0.9844,
+      "step": 6725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.660990893578098,
+      "learning_rate": 4.2894544852958615e-05,
+      "loss": 1.4727,
+      "step": 6726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0514058061248117,
+      "learning_rate": 4.2891632208077576e-05,
+      "loss": 0.7549,
+      "step": 6727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8619631841752833,
+      "learning_rate": 4.288871906527969e-05,
+      "loss": 1.1982,
+      "step": 6728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2975361190463914,
+      "learning_rate": 4.288580542464604e-05,
+      "loss": 0.9304,
+      "step": 6729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.8411421123433565,
+      "learning_rate": 4.2882891286257696e-05,
+      "loss": 1.2188,
+      "step": 6730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.95461991129993,
+      "learning_rate": 4.2879976650195776e-05,
+      "loss": 1.2439,
+      "step": 6731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.12144812895552,
+      "learning_rate": 4.287706151654138e-05,
+      "loss": 0.3877,
+      "step": 6732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9173762072375276,
+      "learning_rate": 4.287414588537565e-05,
+      "loss": 0.8582,
+      "step": 6733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.171859239002819,
+      "learning_rate": 4.2871229756779706e-05,
+      "loss": 1.6777,
+      "step": 6734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.342729787182139,
+      "learning_rate": 4.2868313130834715e-05,
+      "loss": 1.6104,
+      "step": 6735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9798101694974184,
+      "learning_rate": 4.286539600762185e-05,
+      "loss": 1.5547,
+      "step": 6736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.885205825909399,
+      "learning_rate": 4.2862478387222275e-05,
+      "loss": 1.5801,
+      "step": 6737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9833106894317485,
+      "learning_rate": 4.285956026971721e-05,
+      "loss": 1.5439,
+      "step": 6738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.295967447194202,
+      "learning_rate": 4.285664165518784e-05,
+      "loss": 0.9175,
+      "step": 6739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4233219869348894,
+      "learning_rate": 4.285372254371542e-05,
+      "loss": 1.0781,
+      "step": 6740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4452504954738052,
+      "learning_rate": 4.285080293538116e-05,
+      "loss": 1.1611,
+      "step": 6741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.560786320287978,
+      "learning_rate": 4.2847882830266314e-05,
+      "loss": 1.1807,
+      "step": 6742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.7348471208310885,
+      "learning_rate": 4.284496222845217e-05,
+      "loss": 1.0029,
+      "step": 6743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.239895497219133,
+      "learning_rate": 4.2842041130019984e-05,
+      "loss": 1.2695,
+      "step": 6744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.45586897828591,
+      "learning_rate": 4.283911953505106e-05,
+      "loss": 1.0186,
+      "step": 6745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9939888494534936,
+      "learning_rate": 4.2836197443626696e-05,
+      "loss": 1.0615,
+      "step": 6746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.301118066610681,
+      "learning_rate": 4.283327485582822e-05,
+      "loss": 1.3418,
+      "step": 6747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.066227104639603,
+      "learning_rate": 4.2830351771736965e-05,
+      "loss": 1.0244,
+      "step": 6748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.143960623222979,
+      "learning_rate": 4.2827428191434284e-05,
+      "loss": 1.1309,
+      "step": 6749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.473690263253951,
+      "learning_rate": 4.282450411500153e-05,
+      "loss": 1.249,
+      "step": 6750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0684305494808424,
+      "learning_rate": 4.282157954252008e-05,
+      "loss": 1.165,
+      "step": 6751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6829277000510414,
+      "learning_rate": 4.281865447407133e-05,
+      "loss": 1.3728,
+      "step": 6752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.272657780922279,
+      "learning_rate": 4.281572890973668e-05,
+      "loss": 1.0371,
+      "step": 6753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.229485236485824,
+      "learning_rate": 4.281280284959754e-05,
+      "loss": 0.7378,
+      "step": 6754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.340836293888367,
+      "learning_rate": 4.2809876293735365e-05,
+      "loss": 1.0215,
+      "step": 6755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.549474438506714,
+      "learning_rate": 4.280694924223157e-05,
+      "loss": 0.8945,
+      "step": 6756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.331979874173458,
+      "learning_rate": 4.280402169516763e-05,
+      "loss": 1.1699,
+      "step": 6757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.476645583734885,
+      "learning_rate": 4.2801093652625015e-05,
+      "loss": 0.4376,
+      "step": 6758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.364823070142071,
+      "learning_rate": 4.2798165114685216e-05,
+      "loss": 1.3984,
+      "step": 6759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8672558121574796,
+      "learning_rate": 4.2795236081429714e-05,
+      "loss": 1.3506,
+      "step": 6760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5849175876488837,
+      "learning_rate": 4.279230655294004e-05,
+      "loss": 0.7952,
+      "step": 6761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4766094076013863,
+      "learning_rate": 4.2789376529297724e-05,
+      "loss": 1.043,
+      "step": 6762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.398610857436061,
+      "learning_rate": 4.27864460105843e-05,
+      "loss": 0.8291,
+      "step": 6763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6399469351492755,
+      "learning_rate": 4.278351499688132e-05,
+      "loss": 1.7891,
+      "step": 6764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4850137717097582,
+      "learning_rate": 4.278058348827036e-05,
+      "loss": 1.0068,
+      "step": 6765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.600248439693301,
+      "learning_rate": 4.2777651484833e-05,
+      "loss": 1.135,
+      "step": 6766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6548091299000793,
+      "learning_rate": 4.277471898665084e-05,
+      "loss": 1.1104,
+      "step": 6767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.306405889294414,
+      "learning_rate": 4.277178599380548e-05,
+      "loss": 0.8936,
+      "step": 6768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4900016167263925,
+      "learning_rate": 4.2768852506378554e-05,
+      "loss": 1.2207,
+      "step": 6769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.21120984484081,
+      "learning_rate": 4.27659185244517e-05,
+      "loss": 0.4966,
+      "step": 6770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.496459000553138,
+      "learning_rate": 4.2762984048106565e-05,
+      "loss": 1.7988,
+      "step": 6771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9342640492230294,
+      "learning_rate": 4.276004907742481e-05,
+      "loss": 1.3477,
+      "step": 6772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4799633065843296,
+      "learning_rate": 4.275711361248812e-05,
+      "loss": 0.7959,
+      "step": 6773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.117861707255592,
+      "learning_rate": 4.275417765337819e-05,
+      "loss": 0.8789,
+      "step": 6774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.644416479778511,
+      "learning_rate": 4.2751241200176725e-05,
+      "loss": 1.0762,
+      "step": 6775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.760426772403911,
+      "learning_rate": 4.2748304252965444e-05,
+      "loss": 1.3652,
+      "step": 6776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.050338130985667,
+      "learning_rate": 4.2745366811826076e-05,
+      "loss": 1.0557,
+      "step": 6777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.161292441256407,
+      "learning_rate": 4.274242887684038e-05,
+      "loss": 0.7253,
+      "step": 6778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.388951604988928,
+      "learning_rate": 4.2739490448090105e-05,
+      "loss": 0.8027,
+      "step": 6779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.207943179940718,
+      "learning_rate": 4.273655152565704e-05,
+      "loss": 1.5928,
+      "step": 6780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.436888316803908,
+      "learning_rate": 4.2733612109622956e-05,
+      "loss": 0.7373,
+      "step": 6781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.870522923743111,
+      "learning_rate": 4.2730672200069674e-05,
+      "loss": 1.2109,
+      "step": 6782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.600651945393051,
+      "learning_rate": 4.2727731797078995e-05,
+      "loss": 1.3496,
+      "step": 6783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.576524549844408,
+      "learning_rate": 4.272479090073277e-05,
+      "loss": 0.9336,
+      "step": 6784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6820894825605945,
+      "learning_rate": 4.272184951111282e-05,
+      "loss": 0.6941,
+      "step": 6785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.530218161521141,
+      "learning_rate": 4.271890762830102e-05,
+      "loss": 0.9727,
+      "step": 6786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.972813846009086,
+      "learning_rate": 4.271596525237923e-05,
+      "loss": 1.1289,
+      "step": 6787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.967662452560892,
+      "learning_rate": 4.2713022383429334e-05,
+      "loss": 1.5625,
+      "step": 6788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.690915352610612,
+      "learning_rate": 4.271007902153324e-05,
+      "loss": 0.9043,
+      "step": 6789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.632025535987184,
+      "learning_rate": 4.270713516677286e-05,
+      "loss": 0.8896,
+      "step": 6790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7024553554607094,
+      "learning_rate": 4.270419081923012e-05,
+      "loss": 0.8608,
+      "step": 6791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.421023505910904,
+      "learning_rate": 4.2701245978986945e-05,
+      "loss": 1.6816,
+      "step": 6792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5240071601463843,
+      "learning_rate": 4.269830064612531e-05,
+      "loss": 1.5488,
+      "step": 6793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.49721276358441,
+      "learning_rate": 4.269535482072717e-05,
+      "loss": 0.7247,
+      "step": 6794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.463415627758162,
+      "learning_rate": 4.269240850287451e-05,
+      "loss": 1.3271,
+      "step": 6795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.657092946192147,
+      "learning_rate": 4.2689461692649326e-05,
+      "loss": 0.9453,
+      "step": 6796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.604757379458345,
+      "learning_rate": 4.2686514390133615e-05,
+      "loss": 1.4434,
+      "step": 6797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.488908865210808,
+      "learning_rate": 4.2683566595409416e-05,
+      "loss": 1.0098,
+      "step": 6798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.274093061490522,
+      "learning_rate": 4.268061830855876e-05,
+      "loss": 1.6387,
+      "step": 6799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1076353458191113,
+      "learning_rate": 4.267766952966369e-05,
+      "loss": 1.5293,
+      "step": 6800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.65477518486433,
+      "learning_rate": 4.2674720258806277e-05,
+      "loss": 1.4473,
+      "step": 6801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.796814466708176,
+      "learning_rate": 4.267177049606859e-05,
+      "loss": 1.1079,
+      "step": 6802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7036697632571958,
+      "learning_rate": 4.266882024153273e-05,
+      "loss": 0.8726,
+      "step": 6803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.9401458218322665,
+      "learning_rate": 4.266586949528079e-05,
+      "loss": 1.042,
+      "step": 6804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7207357236246996,
+      "learning_rate": 4.26629182573949e-05,
+      "loss": 1.541,
+      "step": 6805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.222837648777235,
+      "learning_rate": 4.2659966527957176e-05,
+      "loss": 1.25,
+      "step": 6806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.414162022500231,
+      "learning_rate": 4.265701430704978e-05,
+      "loss": 1.3965,
+      "step": 6807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.291163915337482,
+      "learning_rate": 4.2654061594754865e-05,
+      "loss": 0.9102,
+      "step": 6808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8710653346697392,
+      "learning_rate": 4.26511083911546e-05,
+      "loss": 0.9863,
+      "step": 6809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.033938462728412,
+      "learning_rate": 4.264815469633118e-05,
+      "loss": 0.9414,
+      "step": 6810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1877913144822285,
+      "learning_rate": 4.264520051036679e-05,
+      "loss": 1.2695,
+      "step": 6811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9493642606901456,
+      "learning_rate": 4.2642245833343654e-05,
+      "loss": 1.2744,
+      "step": 6812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.642093137458098,
+      "learning_rate": 4.2639290665344003e-05,
+      "loss": 0.5361,
+      "step": 6813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1265062898065943,
+      "learning_rate": 4.263633500645007e-05,
+      "loss": 1.3848,
+      "step": 6814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.534382612449794,
+      "learning_rate": 4.2633378856744116e-05,
+      "loss": 1.1416,
+      "step": 6815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0868915097246385,
+      "learning_rate": 4.263042221630841e-05,
+      "loss": 1.2139,
+      "step": 6816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.239115167052522,
+      "learning_rate": 4.2627465085225225e-05,
+      "loss": 0.9038,
+      "step": 6817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.0029023330003195,
+      "learning_rate": 4.262450746357687e-05,
+      "loss": 0.915,
+      "step": 6818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4252940009529462,
+      "learning_rate": 4.262154935144564e-05,
+      "loss": 0.6821,
+      "step": 6819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.476789570453062,
+      "learning_rate": 4.261859074891387e-05,
+      "loss": 1.5234,
+      "step": 6820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.803836066950979,
+      "learning_rate": 4.2615631656063884e-05,
+      "loss": 0.874,
+      "step": 6821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8681431147089986,
+      "learning_rate": 4.2612672072978044e-05,
+      "loss": 0.8999,
+      "step": 6822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.0057978461195525,
+      "learning_rate": 4.260971199973871e-05,
+      "loss": 1.3379,
+      "step": 6823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7055675958900336,
+      "learning_rate": 4.260675143642827e-05,
+      "loss": 1.4082,
+      "step": 6824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.179664968649664,
+      "learning_rate": 4.26037903831291e-05,
+      "loss": 0.9907,
+      "step": 6825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7728566970574333,
+      "learning_rate": 4.260082883992361e-05,
+      "loss": 0.4913,
+      "step": 6826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.216594456191323,
+      "learning_rate": 4.259786680689422e-05,
+      "loss": 1.1777,
+      "step": 6827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.745781176645279,
+      "learning_rate": 4.259490428412335e-05,
+      "loss": 1.1392,
+      "step": 6828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5703290492462063,
+      "learning_rate": 4.259194127169347e-05,
+      "loss": 0.3022,
+      "step": 6829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2972169997763117,
+      "learning_rate": 4.258897776968702e-05,
+      "loss": 0.8018,
+      "step": 6830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.361991795369756,
+      "learning_rate": 4.258601377818647e-05,
+      "loss": 1.1082,
+      "step": 6831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9350415290890572,
+      "learning_rate": 4.2583049297274334e-05,
+      "loss": 1.125,
+      "step": 6832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9829516786503745,
+      "learning_rate": 4.258008432703309e-05,
+      "loss": 1.0781,
+      "step": 6833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9907167715395406,
+      "learning_rate": 4.2577118867545245e-05,
+      "loss": 1.0352,
+      "step": 6834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.128837038367496,
+      "learning_rate": 4.257415291889334e-05,
+      "loss": 0.6284,
+      "step": 6835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3622353501412148,
+      "learning_rate": 4.257118648115992e-05,
+      "loss": 0.9814,
+      "step": 6836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.469693304794171,
+      "learning_rate": 4.256821955442753e-05,
+      "loss": 1.1677,
+      "step": 6837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.50219344236417,
+      "learning_rate": 4.256525213877875e-05,
+      "loss": 0.9707,
+      "step": 6838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.756986602804721,
+      "learning_rate": 4.2562284234296144e-05,
+      "loss": 1.2168,
+      "step": 6839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.771574528352239,
+      "learning_rate": 4.255931584106232e-05,
+      "loss": 0.8545,
+      "step": 6840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8308694074202143,
+      "learning_rate": 4.2556346959159886e-05,
+      "loss": 1.0723,
+      "step": 6841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5533561477024156,
+      "learning_rate": 4.255337758867146e-05,
+      "loss": 0.501,
+      "step": 6842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9322108273969167,
+      "learning_rate": 4.255040772967969e-05,
+      "loss": 0.7441,
+      "step": 6843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9806090265625276,
+      "learning_rate": 4.254743738226721e-05,
+      "loss": 1.1465,
+      "step": 6844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.994321965435687,
+      "learning_rate": 4.2544466546516695e-05,
+      "loss": 1.627,
+      "step": 6845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.365763165517749,
+      "learning_rate": 4.254149522251081e-05,
+      "loss": 1.9434,
+      "step": 6846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.684709475266039,
+      "learning_rate": 4.2538523410332266e-05,
+      "loss": 0.6362,
+      "step": 6847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5180509254996064,
+      "learning_rate": 4.253555111006375e-05,
+      "loss": 0.9609,
+      "step": 6848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.316971823806482,
+      "learning_rate": 4.253257832178798e-05,
+      "loss": 1.2148,
+      "step": 6849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7524179752334463,
+      "learning_rate": 4.252960504558769e-05,
+      "loss": 1.0469,
+      "step": 6850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8941514299648223,
+      "learning_rate": 4.2526631281545646e-05,
+      "loss": 1.2646,
+      "step": 6851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.0486259980960755,
+      "learning_rate": 4.252365702974458e-05,
+      "loss": 1.3594,
+      "step": 6852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.33972656771729,
+      "learning_rate": 4.2520682290267266e-05,
+      "loss": 1.3457,
+      "step": 6853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8706868686348126,
+      "learning_rate": 4.25177070631965e-05,
+      "loss": 1.1836,
+      "step": 6854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.434175405432685,
+      "learning_rate": 4.2514731348615084e-05,
+      "loss": 1.166,
+      "step": 6855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4520802017655075,
+      "learning_rate": 4.2511755146605816e-05,
+      "loss": 1.1084,
+      "step": 6856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.525819668375475,
+      "learning_rate": 4.250877845725154e-05,
+      "loss": 1.0996,
+      "step": 6857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7184783098240617,
+      "learning_rate": 4.2505801280635085e-05,
+      "loss": 0.8779,
+      "step": 6858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.28305617052965,
+      "learning_rate": 4.25028236168393e-05,
+      "loss": 0.7998,
+      "step": 6859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.143541453365326,
+      "learning_rate": 4.2499845465947063e-05,
+      "loss": 0.9551,
+      "step": 6860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.38096101464114,
+      "learning_rate": 4.249686682804125e-05,
+      "loss": 1.1455,
+      "step": 6861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.827105826344822,
+      "learning_rate": 4.249388770320476e-05,
+      "loss": 0.8135,
+      "step": 6862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.644580324265025,
+      "learning_rate": 4.24909080915205e-05,
+      "loss": 1.1055,
+      "step": 6863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.05235166097084,
+      "learning_rate": 4.248792799307138e-05,
+      "loss": 1.7324,
+      "step": 6864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.979268249739578,
+      "learning_rate": 4.248494740794034e-05,
+      "loss": 1.5605,
+      "step": 6865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.484499716421123,
+      "learning_rate": 4.2481966336210346e-05,
+      "loss": 1.0703,
+      "step": 6866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.046214163889325,
+      "learning_rate": 4.2478984777964336e-05,
+      "loss": 1.6133,
+      "step": 6867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8749795912585578,
+      "learning_rate": 4.24760027332853e-05,
+      "loss": 0.7055,
+      "step": 6868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.890538419011963,
+      "learning_rate": 4.247302020225622e-05,
+      "loss": 0.8242,
+      "step": 6869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4665547560911145,
+      "learning_rate": 4.24700371849601e-05,
+      "loss": 1.25,
+      "step": 6870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9487902128630026,
+      "learning_rate": 4.246705368147996e-05,
+      "loss": 1.1406,
+      "step": 6871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6535146539711905,
+      "learning_rate": 4.246406969189882e-05,
+      "loss": 1.0625,
+      "step": 6872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.867408699876428,
+      "learning_rate": 4.246108521629974e-05,
+      "loss": 0.7979,
+      "step": 6873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.881747735469213,
+      "learning_rate": 4.2458100254765765e-05,
+      "loss": 1.1816,
+      "step": 6874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6575325057588737,
+      "learning_rate": 4.245511480737996e-05,
+      "loss": 0.9072,
+      "step": 6875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.081385299362974,
+      "learning_rate": 4.2452128874225425e-05,
+      "loss": 1.1335,
+      "step": 6876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.008005028414411,
+      "learning_rate": 4.2449142455385236e-05,
+      "loss": 0.8569,
+      "step": 6877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.5429624523010945,
+      "learning_rate": 4.244615555094252e-05,
+      "loss": 1.0234,
+      "step": 6878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.529879823175056,
+      "learning_rate": 4.244316816098041e-05,
+      "loss": 0.6716,
+      "step": 6879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.000035285742193,
+      "learning_rate": 4.244018028558202e-05,
+      "loss": 1.1562,
+      "step": 6880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.439151725154304,
+      "learning_rate": 4.243719192483051e-05,
+      "loss": 0.6818,
+      "step": 6881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.209919703359132,
+      "learning_rate": 4.243420307880905e-05,
+      "loss": 1.29,
+      "step": 6882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.156531532332272,
+      "learning_rate": 4.2431213747600824e-05,
+      "loss": 0.9719,
+      "step": 6883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3481011246865306,
+      "learning_rate": 4.242822393128901e-05,
+      "loss": 0.6533,
+      "step": 6884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6798306529786164,
+      "learning_rate": 4.242523362995682e-05,
+      "loss": 1.4531,
+      "step": 6885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.510621776089677,
+      "learning_rate": 4.242224284368746e-05,
+      "loss": 1.5488,
+      "step": 6886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4895791302463266,
+      "learning_rate": 4.241925157256419e-05,
+      "loss": 0.7612,
+      "step": 6887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9442377147073049,
+      "learning_rate": 4.2416259816670235e-05,
+      "loss": 1.125,
+      "step": 6888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.079181625336759,
+      "learning_rate": 4.241326757608886e-05,
+      "loss": 1.0918,
+      "step": 6889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4680437623445037,
+      "learning_rate": 4.241027485090334e-05,
+      "loss": 0.8809,
+      "step": 6890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3301500060973757,
+      "learning_rate": 4.240728164119695e-05,
+      "loss": 0.9873,
+      "step": 6891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0948411963030464,
+      "learning_rate": 4.2404287947053e-05,
+      "loss": 1.292,
+      "step": 6892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0438490770688538,
+      "learning_rate": 4.240129376855481e-05,
+      "loss": 0.8499,
+      "step": 6893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1937526635683304,
+      "learning_rate": 4.2398299105785695e-05,
+      "loss": 1.1357,
+      "step": 6894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5839582630778635,
+      "learning_rate": 4.239530395882899e-05,
+      "loss": 1.0459,
+      "step": 6895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.026465837342002,
+      "learning_rate": 4.239230832776807e-05,
+      "loss": 1.1191,
+      "step": 6896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.700729577738788,
+      "learning_rate": 4.2389312212686284e-05,
+      "loss": 1.0693,
+      "step": 6897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.635289326080406,
+      "learning_rate": 4.238631561366702e-05,
+      "loss": 1.3506,
+      "step": 6898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.01567929216942,
+      "learning_rate": 4.238331853079367e-05,
+      "loss": 0.5122,
+      "step": 6899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7946315526436791,
+      "learning_rate": 4.238032096414964e-05,
+      "loss": 0.9551,
+      "step": 6900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9862249742036546,
+      "learning_rate": 4.237732291381835e-05,
+      "loss": 0.9238,
+      "step": 6901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9889331041366658,
+      "learning_rate": 4.237432437988324e-05,
+      "loss": 1.1719,
+      "step": 6902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.502030737166311,
+      "learning_rate": 4.237132536242776e-05,
+      "loss": 1.2305,
+      "step": 6903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.449383354435578,
+      "learning_rate": 4.236832586153537e-05,
+      "loss": 0.7949,
+      "step": 6904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.566174640394465,
+      "learning_rate": 4.236532587728953e-05,
+      "loss": 0.9775,
+      "step": 6905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.861034577035787,
+      "learning_rate": 4.236232540977375e-05,
+      "loss": 1.4551,
+      "step": 6906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.150714968725027,
+      "learning_rate": 4.235932445907152e-05,
+      "loss": 1.1719,
+      "step": 6907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3524949703989555,
+      "learning_rate": 4.235632302526635e-05,
+      "loss": 0.6123,
+      "step": 6908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.372006357969894,
+      "learning_rate": 4.235332110844178e-05,
+      "loss": 0.8457,
+      "step": 6909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.785429418109526,
+      "learning_rate": 4.235031870868136e-05,
+      "loss": 2.0918,
+      "step": 6910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.513919382894492,
+      "learning_rate": 4.2347315826068626e-05,
+      "loss": 0.8833,
+      "step": 6911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8107867858680966,
+      "learning_rate": 4.2344312460687146e-05,
+      "loss": 1.1084,
+      "step": 6912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5249400669954416,
+      "learning_rate": 4.2341308612620525e-05,
+      "loss": 1.5117,
+      "step": 6913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8945445895954343,
+      "learning_rate": 4.233830428195234e-05,
+      "loss": 0.8174,
+      "step": 6914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6185910362946045,
+      "learning_rate": 4.23352994687662e-05,
+      "loss": 0.9863,
+      "step": 6915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.956621794275894,
+      "learning_rate": 4.233229417314573e-05,
+      "loss": 1.3105,
+      "step": 6916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.65164980582466,
+      "learning_rate": 4.232928839517458e-05,
+      "loss": 1.1045,
+      "step": 6917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.368952329343584,
+      "learning_rate": 4.232628213493638e-05,
+      "loss": 0.7778,
+      "step": 6918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0244615185875636,
+      "learning_rate": 4.2323275392514804e-05,
+      "loss": 1.4434,
+      "step": 6919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7822459777128845,
+      "learning_rate": 4.232026816799353e-05,
+      "loss": 0.6343,
+      "step": 6920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.303923916805568,
+      "learning_rate": 4.2317260461456244e-05,
+      "loss": 0.7832,
+      "step": 6921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.762229435161145,
+      "learning_rate": 4.231425227298664e-05,
+      "loss": 1.4219,
+      "step": 6922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6149688882235202,
+      "learning_rate": 4.231124360266845e-05,
+      "loss": 0.6299,
+      "step": 6923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.64739938485067,
+      "learning_rate": 4.23082344505854e-05,
+      "loss": 1.3613,
+      "step": 6924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.871541730143656,
+      "learning_rate": 4.230522481682123e-05,
+      "loss": 2.2344,
+      "step": 6925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.765037689779892,
+      "learning_rate": 4.23022147014597e-05,
+      "loss": 1.0288,
+      "step": 6926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.013728835333385,
+      "learning_rate": 4.2299204104584575e-05,
+      "loss": 1.1211,
+      "step": 6927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0065462743355322,
+      "learning_rate": 4.229619302627964e-05,
+      "loss": 0.791,
+      "step": 6928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1682046033408846,
+      "learning_rate": 4.22931814666287e-05,
+      "loss": 1.0996,
+      "step": 6929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.88444543007943,
+      "learning_rate": 4.2290169425715567e-05,
+      "loss": 1.0117,
+      "step": 6930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4130160832222822,
+      "learning_rate": 4.228715690362404e-05,
+      "loss": 1.1504,
+      "step": 6931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3914342769561903,
+      "learning_rate": 4.228414390043799e-05,
+      "loss": 0.8828,
+      "step": 6932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0707050455167595,
+      "learning_rate": 4.2281130416241246e-05,
+      "loss": 0.6205,
+      "step": 6933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9686344657785189,
+      "learning_rate": 4.2278116451117674e-05,
+      "loss": 1.2344,
+      "step": 6934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.835336164293536,
+      "learning_rate": 4.2275102005151156e-05,
+      "loss": 1.6699,
+      "step": 6935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.239616297876649,
+      "learning_rate": 4.2272087078425584e-05,
+      "loss": 1.1904,
+      "step": 6936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.97374328273913,
+      "learning_rate": 4.226907167102485e-05,
+      "loss": 0.9209,
+      "step": 6937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7847846212607363,
+      "learning_rate": 4.22660557830329e-05,
+      "loss": 1.2617,
+      "step": 6938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.552793214640839,
+      "learning_rate": 4.2263039414533633e-05,
+      "loss": 0.8711,
+      "step": 6939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.001744605124564,
+      "learning_rate": 4.226002256561101e-05,
+      "loss": 1.1357,
+      "step": 6940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.187112373269288,
+      "learning_rate": 4.225700523634899e-05,
+      "loss": 0.7588,
+      "step": 6941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.281092758044132,
+      "learning_rate": 4.2253987426831534e-05,
+      "loss": 0.6411,
+      "step": 6942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2246002652565346,
+      "learning_rate": 4.225096913714264e-05,
+      "loss": 1.043,
+      "step": 6943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0888263257895625,
+      "learning_rate": 4.224795036736629e-05,
+      "loss": 1.3555,
+      "step": 6944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.167379983221378,
+      "learning_rate": 4.224493111758651e-05,
+      "loss": 0.5166,
+      "step": 6945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.85843396520504,
+      "learning_rate": 4.224191138788731e-05,
+      "loss": 0.9072,
+      "step": 6946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.415554437433331,
+      "learning_rate": 4.2238891178352736e-05,
+      "loss": 1.1523,
+      "step": 6947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.922866245160248,
+      "learning_rate": 4.223587048906684e-05,
+      "loss": 0.9062,
+      "step": 6948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2600744065762632,
+      "learning_rate": 4.223284932011369e-05,
+      "loss": 0.9814,
+      "step": 6949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.234607927716267,
+      "learning_rate": 4.222982767157735e-05,
+      "loss": 1.3125,
+      "step": 6950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.041453384516282,
+      "learning_rate": 4.2226805543541924e-05,
+      "loss": 0.9971,
+      "step": 6951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.472959656398497,
+      "learning_rate": 4.2223782936091514e-05,
+      "loss": 1.0049,
+      "step": 6952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.505293948664461,
+      "learning_rate": 4.2220759849310235e-05,
+      "loss": 1.5518,
+      "step": 6953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.503661430875853,
+      "learning_rate": 4.221773628328222e-05,
+      "loss": 0.915,
+      "step": 6954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.105162584366671,
+      "learning_rate": 4.221471223809162e-05,
+      "loss": 0.9297,
+      "step": 6955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.191483603351516,
+      "learning_rate": 4.221168771382258e-05,
+      "loss": 0.9336,
+      "step": 6956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.2592876569869755,
+      "learning_rate": 4.2208662710559274e-05,
+      "loss": 1.1807,
+      "step": 6957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.193274261100403,
+      "learning_rate": 4.2205637228385896e-05,
+      "loss": 1.0078,
+      "step": 6958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6050990355660995,
+      "learning_rate": 4.2202611267386636e-05,
+      "loss": 0.9834,
+      "step": 6959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.150183818377007,
+      "learning_rate": 4.219958482764571e-05,
+      "loss": 1.1406,
+      "step": 6960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0986752848294685,
+      "learning_rate": 4.219655790924734e-05,
+      "loss": 1.0898,
+      "step": 6961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6482804963602975,
+      "learning_rate": 4.2193530512275766e-05,
+      "loss": 1.6055,
+      "step": 6962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.55400595803621,
+      "learning_rate": 4.2190502636815234e-05,
+      "loss": 1.584,
+      "step": 6963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8485000328908003,
+      "learning_rate": 4.218747428295001e-05,
+      "loss": 1.2676,
+      "step": 6964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.090225216510375,
+      "learning_rate": 4.218444545076438e-05,
+      "loss": 0.686,
+      "step": 6965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0532533695495667,
+      "learning_rate": 4.2181416140342625e-05,
+      "loss": 1.2617,
+      "step": 6966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.961258553579671,
+      "learning_rate": 4.2178386351769054e-05,
+      "loss": 1.3223,
+      "step": 6967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.832811872977553,
+      "learning_rate": 4.217535608512798e-05,
+      "loss": 0.877,
+      "step": 6968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.300620936457769,
+      "learning_rate": 4.2172325340503746e-05,
+      "loss": 0.8992,
+      "step": 6969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.070086159883713,
+      "learning_rate": 4.2169294117980685e-05,
+      "loss": 0.9785,
+      "step": 6970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.069782462737135,
+      "learning_rate": 4.216626241764315e-05,
+      "loss": 1.0664,
+      "step": 6971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.844072798580754,
+      "learning_rate": 4.2163230239575536e-05,
+      "loss": 1.3359,
+      "step": 6972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.011041442034348,
+      "learning_rate": 4.2160197583862204e-05,
+      "loss": 1.4492,
+      "step": 6973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.451506016384801,
+      "learning_rate": 4.2157164450587564e-05,
+      "loss": 1.3057,
+      "step": 6974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9170030699473013,
+      "learning_rate": 4.215413083983601e-05,
+      "loss": 0.9014,
+      "step": 6975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.922731643608889,
+      "learning_rate": 4.215109675169199e-05,
+      "loss": 1.2783,
+      "step": 6976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.52210435365904,
+      "learning_rate": 4.2148062186239926e-05,
+      "loss": 0.8113,
+      "step": 6977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9893026364145863,
+      "learning_rate": 4.214502714356427e-05,
+      "loss": 0.9707,
+      "step": 6978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.913175761109926,
+      "learning_rate": 4.2141991623749485e-05,
+      "loss": 1.8359,
+      "step": 6979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1738277959694834,
+      "learning_rate": 4.213895562688005e-05,
+      "loss": 1.3672,
+      "step": 6980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4489167679786132,
+      "learning_rate": 4.213591915304047e-05,
+      "loss": 0.9414,
+      "step": 6981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.916401187985327,
+      "learning_rate": 4.213288220231523e-05,
+      "loss": 1.6152,
+      "step": 6982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.285673431928373,
+      "learning_rate": 4.2129844774788855e-05,
+      "loss": 1.082,
+      "step": 6983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4611513075214217,
+      "learning_rate": 4.212680687054587e-05,
+      "loss": 0.9971,
+      "step": 6984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9357159655330496,
+      "learning_rate": 4.212376848967082e-05,
+      "loss": 0.939,
+      "step": 6985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8950282196619708,
+      "learning_rate": 4.2120729632248266e-05,
+      "loss": 0.9683,
+      "step": 6986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.959232644127426,
+      "learning_rate": 4.211769029836278e-05,
+      "loss": 0.876,
+      "step": 6987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6425296892724766,
+      "learning_rate": 4.2114650488098936e-05,
+      "loss": 0.8389,
+      "step": 6988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5455715120135087,
+      "learning_rate": 4.211161020154134e-05,
+      "loss": 1.1436,
+      "step": 6989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.142676379890162,
+      "learning_rate": 4.21085694387746e-05,
+      "loss": 0.9707,
+      "step": 6990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9957412054745687,
+      "learning_rate": 4.210552819988334e-05,
+      "loss": 1.2402,
+      "step": 6991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.349090162246451,
+      "learning_rate": 4.210248648495219e-05,
+      "loss": 0.7505,
+      "step": 6992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.081569080970514,
+      "learning_rate": 4.2099444294065796e-05,
+      "loss": 1.3691,
+      "step": 6993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.450187276264139,
+      "learning_rate": 4.2096401627308835e-05,
+      "loss": 0.6479,
+      "step": 6994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.1434809210453,
+      "learning_rate": 4.209335848476598e-05,
+      "loss": 0.6958,
+      "step": 6995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9024691034819785,
+      "learning_rate": 4.2090314866521906e-05,
+      "loss": 1.3945,
+      "step": 6996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7934414109677446,
+      "learning_rate": 4.208727077266134e-05,
+      "loss": 1.1929,
+      "step": 6997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.775639631510623,
+      "learning_rate": 4.208422620326897e-05,
+      "loss": 1.1602,
+      "step": 6998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.562151908492383,
+      "learning_rate": 4.208118115842956e-05,
+      "loss": 0.9329,
+      "step": 6999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.2466025448898135,
+      "learning_rate": 4.2078135638227815e-05,
+      "loss": 1.6621,
+      "step": 7000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.632370083559736,
+      "learning_rate": 4.207508964274852e-05,
+      "loss": 1.3398,
+      "step": 7001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.002332281168384,
+      "learning_rate": 4.207204317207642e-05,
+      "loss": 1.1562,
+      "step": 7002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5437376204805684,
+      "learning_rate": 4.206899622629631e-05,
+      "loss": 1.0366,
+      "step": 7003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.597253618136351,
+      "learning_rate": 4.206594880549299e-05,
+      "loss": 1.5801,
+      "step": 7004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0491846463517507,
+      "learning_rate": 4.2062900909751265e-05,
+      "loss": 0.9248,
+      "step": 7005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.15883011433767,
+      "learning_rate": 4.205985253915595e-05,
+      "loss": 1.1738,
+      "step": 7006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.344932737578489,
+      "learning_rate": 4.205680369379188e-05,
+      "loss": 1.2031,
+      "step": 7007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.398190165636712,
+      "learning_rate": 4.2053754373743904e-05,
+      "loss": 1.2109,
+      "step": 7008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.776227319978538,
+      "learning_rate": 4.2050704579096895e-05,
+      "loss": 1.2529,
+      "step": 7009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.003228861620069,
+      "learning_rate": 4.204765430993571e-05,
+      "loss": 0.8691,
+      "step": 7010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.184039835650884,
+      "learning_rate": 4.204460356634525e-05,
+      "loss": 1.5801,
+      "step": 7011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9048217676042847,
+      "learning_rate": 4.204155234841042e-05,
+      "loss": 1.269,
+      "step": 7012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.584250705550842,
+      "learning_rate": 4.203850065621612e-05,
+      "loss": 1.3965,
+      "step": 7013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0578124313861292,
+      "learning_rate": 4.2035448489847284e-05,
+      "loss": 1.1543,
+      "step": 7014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.863958801623204,
+      "learning_rate": 4.203239584938885e-05,
+      "loss": 1.834,
+      "step": 7015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.052141850835985,
+      "learning_rate": 4.202934273492577e-05,
+      "loss": 0.8584,
+      "step": 7016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2498589998616763,
+      "learning_rate": 4.202628914654302e-05,
+      "loss": 1.2031,
+      "step": 7017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5759138579863428,
+      "learning_rate": 4.202323508432557e-05,
+      "loss": 1.1445,
+      "step": 7018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5850275751008387,
+      "learning_rate": 4.202018054835842e-05,
+      "loss": 1.4746,
+      "step": 7019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.2081384645170425,
+      "learning_rate": 4.201712553872658e-05,
+      "loss": 1.1514,
+      "step": 7020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4179812725279026,
+      "learning_rate": 4.201407005551505e-05,
+      "loss": 1.2637,
+      "step": 7021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9084406974088775,
+      "learning_rate": 4.201101409880889e-05,
+      "loss": 0.9316,
+      "step": 7022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.102863715185114,
+      "learning_rate": 4.200795766869312e-05,
+      "loss": 1.0234,
+      "step": 7023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3491680336423073,
+      "learning_rate": 4.200490076525282e-05,
+      "loss": 0.36,
+      "step": 7024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.029343265001978,
+      "learning_rate": 4.200184338857305e-05,
+      "loss": 1.2441,
+      "step": 7025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.447864946098898,
+      "learning_rate": 4.199878553873891e-05,
+      "loss": 1.3623,
+      "step": 7026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.538283477998262,
+      "learning_rate": 4.1995727215835476e-05,
+      "loss": 0.7393,
+      "step": 7027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1302599041827963,
+      "learning_rate": 4.1992668419947884e-05,
+      "loss": 0.8873,
+      "step": 7028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.803383731526857,
+      "learning_rate": 4.198960915116125e-05,
+      "loss": 1.3105,
+      "step": 7029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.056371203288376,
+      "learning_rate": 4.1986549409560694e-05,
+      "loss": 1.5117,
+      "step": 7030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.085519143519808,
+      "learning_rate": 4.198348919523139e-05,
+      "loss": 1.8262,
+      "step": 7031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2913663031513067,
+      "learning_rate": 4.198042850825849e-05,
+      "loss": 1.2305,
+      "step": 7032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3630417379276203,
+      "learning_rate": 4.197736734872719e-05,
+      "loss": 1.5957,
+      "step": 7033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3398457488115363,
+      "learning_rate": 4.197430571672267e-05,
+      "loss": 1.2168,
+      "step": 7034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.045640431221569,
+      "learning_rate": 4.1971243612330126e-05,
+      "loss": 0.8018,
+      "step": 7035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5060063688436256,
+      "learning_rate": 4.1968181035634775e-05,
+      "loss": 0.6992,
+      "step": 7036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.3743076321564525,
+      "learning_rate": 4.1965117986721856e-05,
+      "loss": 1.207,
+      "step": 7037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4669715177999383,
+      "learning_rate": 4.1962054465676614e-05,
+      "loss": 0.7109,
+      "step": 7038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.913464267213937,
+      "learning_rate": 4.19589904725843e-05,
+      "loss": 1.7207,
+      "step": 7039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1520957180773457,
+      "learning_rate": 4.195592600753019e-05,
+      "loss": 1.8906,
+      "step": 7040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5831436784919477,
+      "learning_rate": 4.1952861070599555e-05,
+      "loss": 0.8604,
+      "step": 7041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.938239146638155,
+      "learning_rate": 4.19497956618777e-05,
+      "loss": 0.7632,
+      "step": 7042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0540983212375683,
+      "learning_rate": 4.194672978144993e-05,
+      "loss": 1.0312,
+      "step": 7043
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.86782638541455,
+      "learning_rate": 4.194366342940157e-05,
+      "loss": 0.75,
+      "step": 7044
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0960219852590307,
+      "learning_rate": 4.194059660581796e-05,
+      "loss": 1.0938,
+      "step": 7045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.945558831118771,
+      "learning_rate": 4.193752931078444e-05,
+      "loss": 0.7085,
+      "step": 7046
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.648859243094221,
+      "learning_rate": 4.193446154438637e-05,
+      "loss": 1.3926,
+      "step": 7047
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0837262609602143,
+      "learning_rate": 4.193139330670913e-05,
+      "loss": 1.1641,
+      "step": 7048
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.104794810808791,
+      "learning_rate": 4.1928324597838106e-05,
+      "loss": 1.5391,
+      "step": 7049
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.950968033717714,
+      "learning_rate": 4.1925255417858714e-05,
+      "loss": 2.041,
+      "step": 7050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.640603420914278,
+      "learning_rate": 4.1922185766856334e-05,
+      "loss": 1.1465,
+      "step": 7051
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.024819858956593,
+      "learning_rate": 4.191911564491643e-05,
+      "loss": 1.2617,
+      "step": 7052
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5690944765507084,
+      "learning_rate": 4.191604505212442e-05,
+      "loss": 1.0713,
+      "step": 7053
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.862783779815309,
+      "learning_rate": 4.191297398856576e-05,
+      "loss": 1.5781,
+      "step": 7054
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.172359055840603,
+      "learning_rate": 4.1909902454325924e-05,
+      "loss": 0.4911,
+      "step": 7055
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2751284468724102,
+      "learning_rate": 4.190683044949038e-05,
+      "loss": 0.7129,
+      "step": 7056
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.588391171929801,
+      "learning_rate": 4.1903757974144626e-05,
+      "loss": 0.6875,
+      "step": 7057
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0714801302727346,
+      "learning_rate": 4.1900685028374176e-05,
+      "loss": 1.1011,
+      "step": 7058
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.375476594949129,
+      "learning_rate": 4.189761161226454e-05,
+      "loss": 1.4741,
+      "step": 7059
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7519513056405893,
+      "learning_rate": 4.189453772590125e-05,
+      "loss": 1.2852,
+      "step": 7060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8689945357740347,
+      "learning_rate": 4.1891463369369854e-05,
+      "loss": 1.2324,
+      "step": 7061
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4156189872795175,
+      "learning_rate": 4.188838854275591e-05,
+      "loss": 0.7031,
+      "step": 7062
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.813212843994355,
+      "learning_rate": 4.188531324614498e-05,
+      "loss": 1.166,
+      "step": 7063
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.577204962049459,
+      "learning_rate": 4.188223747962267e-05,
+      "loss": 0.7969,
+      "step": 7064
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7388709898443975,
+      "learning_rate": 4.1879161243274554e-05,
+      "loss": 1.666,
+      "step": 7065
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.3137133348148415,
+      "learning_rate": 4.1876084537186255e-05,
+      "loss": 1.2207,
+      "step": 7066
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.008754383559321,
+      "learning_rate": 4.187300736144339e-05,
+      "loss": 0.5923,
+      "step": 7067
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.336037413197493,
+      "learning_rate": 4.1869929716131605e-05,
+      "loss": 0.6851,
+      "step": 7068
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1008567334024058,
+      "learning_rate": 4.186685160133654e-05,
+      "loss": 0.7349,
+      "step": 7069
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.749681627744397,
+      "learning_rate": 4.186377301714386e-05,
+      "loss": 1.0352,
+      "step": 7070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.573184527616704,
+      "learning_rate": 4.1860693963639255e-05,
+      "loss": 1.041,
+      "step": 7071
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.046946616594714,
+      "learning_rate": 4.1857614440908386e-05,
+      "loss": 1.668,
+      "step": 7072
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9919458603560947,
+      "learning_rate": 4.185453444903698e-05,
+      "loss": 1.5879,
+      "step": 7073
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.336480801403674,
+      "learning_rate": 4.185145398811073e-05,
+      "loss": 1.0762,
+      "step": 7074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.319909365790961,
+      "learning_rate": 4.1848373058215385e-05,
+      "loss": 1.791,
+      "step": 7075
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.026797652566714,
+      "learning_rate": 4.184529165943667e-05,
+      "loss": 1.1953,
+      "step": 7076
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.149633269501949,
+      "learning_rate": 4.1842209791860354e-05,
+      "loss": 1.6836,
+      "step": 7077
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0099170802661863,
+      "learning_rate": 4.183912745557219e-05,
+      "loss": 1.0293,
+      "step": 7078
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.5134794878822255,
+      "learning_rate": 4.183604465065797e-05,
+      "loss": 1.0693,
+      "step": 7079
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.812682600807724,
+      "learning_rate": 4.183296137720348e-05,
+      "loss": 0.8525,
+      "step": 7080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.488611316198569,
+      "learning_rate": 4.1829877635294525e-05,
+      "loss": 1.0889,
+      "step": 7081
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7866024250527066,
+      "learning_rate": 4.182679342501692e-05,
+      "loss": 0.8825,
+      "step": 7082
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.30110121291202,
+      "learning_rate": 4.182370874645651e-05,
+      "loss": 1.0791,
+      "step": 7083
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.229864943994822,
+      "learning_rate": 4.182062359969914e-05,
+      "loss": 1.1021,
+      "step": 7084
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3771863210190505,
+      "learning_rate": 4.181753798483066e-05,
+      "loss": 0.5332,
+      "step": 7085
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.142654143007314,
+      "learning_rate": 4.1814451901936935e-05,
+      "loss": 0.9082,
+      "step": 7086
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.566379643118508,
+      "learning_rate": 4.181136535110387e-05,
+      "loss": 1.1074,
+      "step": 7087
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.504793157539191,
+      "learning_rate": 4.180827833241735e-05,
+      "loss": 0.8423,
+      "step": 7088
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.490384955948911,
+      "learning_rate": 4.1805190845963284e-05,
+      "loss": 0.7969,
+      "step": 7089
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5984383000999407,
+      "learning_rate": 4.1802102891827596e-05,
+      "loss": 1.0645,
+      "step": 7090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0674357454769967,
+      "learning_rate": 4.1799014470096234e-05,
+      "loss": 0.9302,
+      "step": 7091
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5023964839717916,
+      "learning_rate": 4.1795925580855126e-05,
+      "loss": 1.334,
+      "step": 7092
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8188364493568363,
+      "learning_rate": 4.179283622419026e-05,
+      "loss": 1.127,
+      "step": 7093
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5666195390346225,
+      "learning_rate": 4.178974640018759e-05,
+      "loss": 0.8462,
+      "step": 7094
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1705134433857776,
+      "learning_rate": 4.178665610893311e-05,
+      "loss": 1.168,
+      "step": 7095
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.16616238419847,
+      "learning_rate": 4.178356535051282e-05,
+      "loss": 1.3916,
+      "step": 7096
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.3797823244813054,
+      "learning_rate": 4.178047412501275e-05,
+      "loss": 1.4805,
+      "step": 7097
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.899960005489836,
+      "learning_rate": 4.177738243251892e-05,
+      "loss": 0.9756,
+      "step": 7098
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.368860523909119,
+      "learning_rate": 4.1774290273117354e-05,
+      "loss": 0.8662,
+      "step": 7099
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5396935251022974,
+      "learning_rate": 4.177119764689412e-05,
+      "loss": 0.9243,
+      "step": 7100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9377415334543664,
+      "learning_rate": 4.17681045539353e-05,
+      "loss": 1.2676,
+      "step": 7101
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.426219366195652,
+      "learning_rate": 4.176501099432693e-05,
+      "loss": 1.0469,
+      "step": 7102
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9077266100333934,
+      "learning_rate": 4.176191696815514e-05,
+      "loss": 0.938,
+      "step": 7103
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.066163965846266,
+      "learning_rate": 4.175882247550603e-05,
+      "loss": 1.8652,
+      "step": 7104
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4614447572854976,
+      "learning_rate": 4.175572751646571e-05,
+      "loss": 0.8809,
+      "step": 7105
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.385550650538738,
+      "learning_rate": 4.175263209112032e-05,
+      "loss": 1.438,
+      "step": 7106
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.625194630964378,
+      "learning_rate": 4.174953619955599e-05,
+      "loss": 0.8472,
+      "step": 7107
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.611631039958667,
+      "learning_rate": 4.1746439841858895e-05,
+      "loss": 1.3281,
+      "step": 7108
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7489499211338413,
+      "learning_rate": 4.174334301811519e-05,
+      "loss": 0.9678,
+      "step": 7109
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.525239849973907,
+      "learning_rate": 4.174024572841107e-05,
+      "loss": 1.1006,
+      "step": 7110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.854274014498487,
+      "learning_rate": 4.173714797283272e-05,
+      "loss": 0.8311,
+      "step": 7111
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.012754507117002,
+      "learning_rate": 4.1734049751466365e-05,
+      "loss": 1.3008,
+      "step": 7112
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3855318155495784,
+      "learning_rate": 4.1730951064398204e-05,
+      "loss": 0.8245,
+      "step": 7113
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6946423428975033,
+      "learning_rate": 4.17278519117145e-05,
+      "loss": 1.5293,
+      "step": 7114
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1576634631880784,
+      "learning_rate": 4.1724752293501476e-05,
+      "loss": 1.1162,
+      "step": 7115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.703012717676816,
+      "learning_rate": 4.172165220984541e-05,
+      "loss": 0.7588,
+      "step": 7116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.907808870146003,
+      "learning_rate": 4.171855166083257e-05,
+      "loss": 0.7388,
+      "step": 7117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.846342709579194,
+      "learning_rate": 4.171545064654925e-05,
+      "loss": 1.1406,
+      "step": 7118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.886872921104199,
+      "learning_rate": 4.171234916708173e-05,
+      "loss": 1.1152,
+      "step": 7119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1354199448544082,
+      "learning_rate": 4.170924722251635e-05,
+      "loss": 0.8354,
+      "step": 7120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1710607525492613,
+      "learning_rate": 4.170614481293941e-05,
+      "loss": 1.3672,
+      "step": 7121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.241818729170014,
+      "learning_rate": 4.170304193843727e-05,
+      "loss": 0.9219,
+      "step": 7122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0832156768895174,
+      "learning_rate": 4.169993859909627e-05,
+      "loss": 0.8599,
+      "step": 7123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4976409272077977,
+      "learning_rate": 4.169683479500277e-05,
+      "loss": 0.7977,
+      "step": 7124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0827457967928598,
+      "learning_rate": 4.169373052624316e-05,
+      "loss": 0.876,
+      "step": 7125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0710125384756424,
+      "learning_rate": 4.169062579290383e-05,
+      "loss": 1.125,
+      "step": 7126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.370990550823205,
+      "learning_rate": 4.168752059507117e-05,
+      "loss": 0.7008,
+      "step": 7127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.48570772328317,
+      "learning_rate": 4.168441493283161e-05,
+      "loss": 0.8987,
+      "step": 7128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6389504263768875,
+      "learning_rate": 4.168130880627158e-05,
+      "loss": 1.4492,
+      "step": 7129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.523761362368592,
+      "learning_rate": 4.1678202215477513e-05,
+      "loss": 0.772,
+      "step": 7130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.0281497359718585,
+      "learning_rate": 4.167509516053587e-05,
+      "loss": 0.7266,
+      "step": 7131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5428536150251135,
+      "learning_rate": 4.1671987641533106e-05,
+      "loss": 0.9375,
+      "step": 7132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.602234791488664,
+      "learning_rate": 4.166887965855573e-05,
+      "loss": 2.0332,
+      "step": 7133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.756752653414549,
+      "learning_rate": 4.166577121169021e-05,
+      "loss": 1.2676,
+      "step": 7134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.009609474531349,
+      "learning_rate": 4.166266230102306e-05,
+      "loss": 1.1318,
+      "step": 7135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.36389333821929,
+      "learning_rate": 4.16595529266408e-05,
+      "loss": 0.8872,
+      "step": 7136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2174929429278496,
+      "learning_rate": 4.1656443088629964e-05,
+      "loss": 0.7686,
+      "step": 7137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3503261623890834,
+      "learning_rate": 4.1653332787077107e-05,
+      "loss": 0.8359,
+      "step": 7138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.434769394236546,
+      "learning_rate": 4.1650222022068776e-05,
+      "loss": 1.2812,
+      "step": 7139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.147218984310605,
+      "learning_rate": 4.164711079369153e-05,
+      "loss": 1.1797,
+      "step": 7140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.469949067518175,
+      "learning_rate": 4.164399910203198e-05,
+      "loss": 0.8804,
+      "step": 7141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.666434792618786,
+      "learning_rate": 4.1640886947176716e-05,
+      "loss": 1.6426,
+      "step": 7142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.318241620949863,
+      "learning_rate": 4.163777432921233e-05,
+      "loss": 1.0361,
+      "step": 7143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.725619205484716,
+      "learning_rate": 4.163466124822547e-05,
+      "loss": 1.8481,
+      "step": 7144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.730192344900044,
+      "learning_rate": 4.163154770430275e-05,
+      "loss": 1.1761,
+      "step": 7145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.678377557229981,
+      "learning_rate": 4.162843369753083e-05,
+      "loss": 1.0283,
+      "step": 7146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.461909894634462,
+      "learning_rate": 4.1625319227996375e-05,
+      "loss": 1.4609,
+      "step": 7147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.342398453811561,
+      "learning_rate": 4.162220429578605e-05,
+      "loss": 0.8672,
+      "step": 7148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.700352034754503,
+      "learning_rate": 4.1619088900986544e-05,
+      "loss": 1.1904,
+      "step": 7149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.644627443984193,
+      "learning_rate": 4.161597304368456e-05,
+      "loss": 0.6973,
+      "step": 7150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3687879735474526,
+      "learning_rate": 4.1612856723966814e-05,
+      "loss": 1.3408,
+      "step": 7151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7361095792975556,
+      "learning_rate": 4.160973994192003e-05,
+      "loss": 0.9561,
+      "step": 7152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2437973989209903,
+      "learning_rate": 4.160662269763094e-05,
+      "loss": 0.9082,
+      "step": 7153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.306423605643113,
+      "learning_rate": 4.1603504991186306e-05,
+      "loss": 0.9902,
+      "step": 7154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8681090326585865,
+      "learning_rate": 4.160038682267288e-05,
+      "loss": 1.1289,
+      "step": 7155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.520451435343924,
+      "learning_rate": 4.159726819217745e-05,
+      "loss": 1.1777,
+      "step": 7156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.007496960805977,
+      "learning_rate": 4.1594149099786806e-05,
+      "loss": 0.918,
+      "step": 7157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.913120475459437,
+      "learning_rate": 4.1591029545587746e-05,
+      "loss": 1.1738,
+      "step": 7158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.333610501350692,
+      "learning_rate": 4.158790952966709e-05,
+      "loss": 1.3574,
+      "step": 7159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.352911673353884,
+      "learning_rate": 4.158478905211166e-05,
+      "loss": 1.543,
+      "step": 7160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.586588958952418,
+      "learning_rate": 4.158166811300831e-05,
+      "loss": 1.2412,
+      "step": 7161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9078304136483912,
+      "learning_rate": 4.1578546712443885e-05,
+      "loss": 1.3008,
+      "step": 7162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.754061818815122,
+      "learning_rate": 4.1575424850505255e-05,
+      "loss": 1.5869,
+      "step": 7163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.892805936652818,
+      "learning_rate": 4.1572302527279284e-05,
+      "loss": 1.0527,
+      "step": 7164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.842987293624258,
+      "learning_rate": 4.156917974285289e-05,
+      "loss": 1.1372,
+      "step": 7165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6822511737816295,
+      "learning_rate": 4.156605649731298e-05,
+      "loss": 1.0156,
+      "step": 7166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3629440697610358,
+      "learning_rate": 4.156293279074645e-05,
+      "loss": 1.0785,
+      "step": 7167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8174445353444337,
+      "learning_rate": 4.1559808623240246e-05,
+      "loss": 0.5376,
+      "step": 7168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.47754010923951,
+      "learning_rate": 4.155668399488131e-05,
+      "loss": 1.5273,
+      "step": 7169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.147744330992246,
+      "learning_rate": 4.1553558905756596e-05,
+      "loss": 1.0762,
+      "step": 7170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.444984974727216,
+      "learning_rate": 4.155043335595308e-05,
+      "loss": 0.8333,
+      "step": 7171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.708011154809797,
+      "learning_rate": 4.154730734555774e-05,
+      "loss": 1.2734,
+      "step": 7172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5655504074125797,
+      "learning_rate": 4.154418087465757e-05,
+      "loss": 0.989,
+      "step": 7173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.706935776297079,
+      "learning_rate": 4.154105394333958e-05,
+      "loss": 0.489,
+      "step": 7174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.400659040571786,
+      "learning_rate": 4.15379265516908e-05,
+      "loss": 1.0791,
+      "step": 7175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.47839167662934,
+      "learning_rate": 4.1534798699798254e-05,
+      "loss": 1.6992,
+      "step": 7176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.226610234163781,
+      "learning_rate": 4.1531670387748995e-05,
+      "loss": 0.8701,
+      "step": 7177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.1470869779277555,
+      "learning_rate": 4.152854161563007e-05,
+      "loss": 1.1758,
+      "step": 7178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.238130712692355,
+      "learning_rate": 4.152541238352857e-05,
+      "loss": 1.209,
+      "step": 7179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.1895639381163665,
+      "learning_rate": 4.152228269153157e-05,
+      "loss": 0.834,
+      "step": 7180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6891333473709755,
+      "learning_rate": 4.1519152539726164e-05,
+      "loss": 0.834,
+      "step": 7181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.72388080833662,
+      "learning_rate": 4.1516021928199464e-05,
+      "loss": 0.9292,
+      "step": 7182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6328778660953938,
+      "learning_rate": 4.151289085703861e-05,
+      "loss": 1.2715,
+      "step": 7183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8347961841122196,
+      "learning_rate": 4.150975932633071e-05,
+      "loss": 0.75,
+      "step": 7184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.148626233421588,
+      "learning_rate": 4.150662733616294e-05,
+      "loss": 1.126,
+      "step": 7185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9000836338954827,
+      "learning_rate": 4.150349488662245e-05,
+      "loss": 0.6919,
+      "step": 7186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6280659527217862,
+      "learning_rate": 4.1500361977796416e-05,
+      "loss": 1.3594,
+      "step": 7187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.678801930119675,
+      "learning_rate": 4.149722860977202e-05,
+      "loss": 0.3926,
+      "step": 7188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.189051082428039,
+      "learning_rate": 4.149409478263647e-05,
+      "loss": 1.3672,
+      "step": 7189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.790877011971961,
+      "learning_rate": 4.149096049647698e-05,
+      "loss": 0.7446,
+      "step": 7190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.665176189814123,
+      "learning_rate": 4.1487825751380784e-05,
+      "loss": 0.9973,
+      "step": 7191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4829105883572633,
+      "learning_rate": 4.148469054743509e-05,
+      "loss": 0.4834,
+      "step": 7192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0844874681959396,
+      "learning_rate": 4.1481554884727174e-05,
+      "loss": 0.8828,
+      "step": 7193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.420199528361119,
+      "learning_rate": 4.14784187633443e-05,
+      "loss": 0.9849,
+      "step": 7194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.759452649871326,
+      "learning_rate": 4.147528218337374e-05,
+      "loss": 1.0518,
+      "step": 7195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.805513988056844,
+      "learning_rate": 4.147214514490278e-05,
+      "loss": 0.9038,
+      "step": 7196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.7630556948874085,
+      "learning_rate": 4.1469007648018726e-05,
+      "loss": 1.418,
+      "step": 7197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.237737130250934,
+      "learning_rate": 4.14658696928089e-05,
+      "loss": 1.498,
+      "step": 7198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.039746460843243,
+      "learning_rate": 4.146273127936062e-05,
+      "loss": 1.5625,
+      "step": 7199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.647659251427738,
+      "learning_rate": 4.145959240776123e-05,
+      "loss": 0.6957,
+      "step": 7200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.09431681808946,
+      "learning_rate": 4.145645307809808e-05,
+      "loss": 1.1299,
+      "step": 7201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.431777039815091,
+      "learning_rate": 4.145331329045855e-05,
+      "loss": 0.9795,
+      "step": 7202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1892424663954997,
+      "learning_rate": 4.145017304492999e-05,
+      "loss": 1.0332,
+      "step": 7203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9800226736215472,
+      "learning_rate": 4.1447032341599835e-05,
+      "loss": 0.8313,
+      "step": 7204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3622730973691324,
+      "learning_rate": 4.1443891180555454e-05,
+      "loss": 1.0479,
+      "step": 7205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.900428379562613,
+      "learning_rate": 4.144074956188427e-05,
+      "loss": 1.1152,
+      "step": 7206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.926348520451208,
+      "learning_rate": 4.143760748567373e-05,
+      "loss": 1.5762,
+      "step": 7207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.019218512266322,
+      "learning_rate": 4.143446495201126e-05,
+      "loss": 0.7996,
+      "step": 7208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1462453304677656,
+      "learning_rate": 4.1431321960984325e-05,
+      "loss": 1.4102,
+      "step": 7209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.625503306453547,
+      "learning_rate": 4.1428178512680383e-05,
+      "loss": 1.1455,
+      "step": 7210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.642074367778899,
+      "learning_rate": 4.1425034607186925e-05,
+      "loss": 0.9014,
+      "step": 7211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.776351585592835,
+      "learning_rate": 4.142189024459144e-05,
+      "loss": 1.0103,
+      "step": 7212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6976339322163176,
+      "learning_rate": 4.1418745424981434e-05,
+      "loss": 1.418,
+      "step": 7213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9112492423759355,
+      "learning_rate": 4.1415600148444425e-05,
+      "loss": 0.9062,
+      "step": 7214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0622026633745536,
+      "learning_rate": 4.141245441506795e-05,
+      "loss": 0.7715,
+      "step": 7215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.052106698807124,
+      "learning_rate": 4.1409308224939545e-05,
+      "loss": 1.4648,
+      "step": 7216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2134656706087354,
+      "learning_rate": 4.140616157814678e-05,
+      "loss": 1.167,
+      "step": 7217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8760110487108927,
+      "learning_rate": 4.1403014474777215e-05,
+      "loss": 1.0557,
+      "step": 7218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.243475159631259,
+      "learning_rate": 4.1399866914918433e-05,
+      "loss": 1.2207,
+      "step": 7219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.631297015100883,
+      "learning_rate": 4.1396718898658025e-05,
+      "loss": 0.8921,
+      "step": 7220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.218381060256559,
+      "learning_rate": 4.1393570426083614e-05,
+      "loss": 1.2422,
+      "step": 7221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.250718550844068,
+      "learning_rate": 4.139042149728281e-05,
+      "loss": 1.1816,
+      "step": 7222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.656676612102471,
+      "learning_rate": 4.138727211234325e-05,
+      "loss": 0.8906,
+      "step": 7223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.317943951566772,
+      "learning_rate": 4.138412227135258e-05,
+      "loss": 0.8184,
+      "step": 7224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.561328035828738,
+      "learning_rate": 4.138097197439845e-05,
+      "loss": 0.4854,
+      "step": 7225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6019051357689373,
+      "learning_rate": 4.137782122156854e-05,
+      "loss": 1.1689,
+      "step": 7226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.222752685574929,
+      "learning_rate": 4.137467001295053e-05,
+      "loss": 0.9385,
+      "step": 7227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.616043615170085,
+      "learning_rate": 4.137151834863213e-05,
+      "loss": 0.7051,
+      "step": 7228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.701507034440077,
+      "learning_rate": 4.136836622870103e-05,
+      "loss": 1.207,
+      "step": 7229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5410033708161515,
+      "learning_rate": 4.136521365324497e-05,
+      "loss": 1.2695,
+      "step": 7230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3433398078550463,
+      "learning_rate": 4.136206062235167e-05,
+      "loss": 1.3633,
+      "step": 7231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.194148237760275,
+      "learning_rate": 4.1358907136108885e-05,
+      "loss": 1.0513,
+      "step": 7232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.207734393629619,
+      "learning_rate": 4.1355753194604376e-05,
+      "loss": 1.8789,
+      "step": 7233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8185163813249907,
+      "learning_rate": 4.135259879792591e-05,
+      "loss": 1.4902,
+      "step": 7234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8696448464161173,
+      "learning_rate": 4.134944394616128e-05,
+      "loss": 0.5406,
+      "step": 7235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.110937444820049,
+      "learning_rate": 4.134628863939828e-05,
+      "loss": 0.6064,
+      "step": 7236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.38144522582197,
+      "learning_rate": 4.134313287772472e-05,
+      "loss": 1.2852,
+      "step": 7237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.354520417437666,
+      "learning_rate": 4.1339976661228425e-05,
+      "loss": 0.9023,
+      "step": 7238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.716611368488335,
+      "learning_rate": 4.1336819989997234e-05,
+      "loss": 1.5195,
+      "step": 7239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7490840285821427,
+      "learning_rate": 4.1333662864118994e-05,
+      "loss": 0.7305,
+      "step": 7240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.061982292165858,
+      "learning_rate": 4.133050528368157e-05,
+      "loss": 0.7443,
+      "step": 7241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2550187827806134,
+      "learning_rate": 4.132734724877282e-05,
+      "loss": 1.0735,
+      "step": 7242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.079101447826148,
+      "learning_rate": 4.132418875948065e-05,
+      "loss": 1.0869,
+      "step": 7243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.722493598371119,
+      "learning_rate": 4.132102981589295e-05,
+      "loss": 0.7881,
+      "step": 7244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3004307633750942,
+      "learning_rate": 4.1317870418097635e-05,
+      "loss": 0.7324,
+      "step": 7245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6746201662748823,
+      "learning_rate": 4.131471056618263e-05,
+      "loss": 1.0117,
+      "step": 7246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7933463741656626,
+      "learning_rate": 4.131155026023587e-05,
+      "loss": 1.1875,
+      "step": 7247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.286547109174798,
+      "learning_rate": 4.130838950034531e-05,
+      "loss": 0.9629,
+      "step": 7248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.593619208517542,
+      "learning_rate": 4.130522828659892e-05,
+      "loss": 1.2129,
+      "step": 7249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.2354209232878715,
+      "learning_rate": 4.1302066619084644e-05,
+      "loss": 1.0825,
+      "step": 7250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7995542371652458,
+      "learning_rate": 4.12989044978905e-05,
+      "loss": 1.1323,
+      "step": 7251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.863409146188606,
+      "learning_rate": 4.129574192310448e-05,
+      "loss": 1.2549,
+      "step": 7252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.329582767603401,
+      "learning_rate": 4.1292578894814596e-05,
+      "loss": 0.52,
+      "step": 7253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.954569394392854,
+      "learning_rate": 4.128941541310888e-05,
+      "loss": 1.3389,
+      "step": 7254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.882551392380527,
+      "learning_rate": 4.1286251478075354e-05,
+      "loss": 1.0986,
+      "step": 7255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.120668384656112,
+      "learning_rate": 4.128308708980208e-05,
+      "loss": 1.2124,
+      "step": 7256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.498840003564672,
+      "learning_rate": 4.127992224837712e-05,
+      "loss": 1.4023,
+      "step": 7257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.516477088656423,
+      "learning_rate": 4.127675695388856e-05,
+      "loss": 0.6943,
+      "step": 7258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0839995274524847,
+      "learning_rate": 4.1273591206424477e-05,
+      "loss": 1.0103,
+      "step": 7259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.185027346139981,
+      "learning_rate": 4.1270425006072974e-05,
+      "loss": 1.3809,
+      "step": 7260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.432101362603324,
+      "learning_rate": 4.126725835292217e-05,
+      "loss": 1.1572,
+      "step": 7261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1260776031412534,
+      "learning_rate": 4.1264091247060184e-05,
+      "loss": 1.3574,
+      "step": 7262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.637342591130922,
+      "learning_rate": 4.126092368857516e-05,
+      "loss": 1.3574,
+      "step": 7263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2875427825762515,
+      "learning_rate": 4.1257755677555255e-05,
+      "loss": 0.5598,
+      "step": 7264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.234107427835201,
+      "learning_rate": 4.125458721408862e-05,
+      "loss": 1.3291,
+      "step": 7265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.552131050338483,
+      "learning_rate": 4.125141829826345e-05,
+      "loss": 1.0811,
+      "step": 7266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0181112406743904,
+      "learning_rate": 4.124824893016792e-05,
+      "loss": 0.3627,
+      "step": 7267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.470990026358357,
+      "learning_rate": 4.124507910989023e-05,
+      "loss": 1.2852,
+      "step": 7268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.089894676512207,
+      "learning_rate": 4.124190883751861e-05,
+      "loss": 0.4778,
+      "step": 7269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2806233629726265,
+      "learning_rate": 4.123873811314128e-05,
+      "loss": 1.0024,
+      "step": 7270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6833899567148451,
+      "learning_rate": 4.123556693684647e-05,
+      "loss": 0.5273,
+      "step": 7271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8125979936270458,
+      "learning_rate": 4.123239530872245e-05,
+      "loss": 1.2612,
+      "step": 7272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.122494398453401,
+      "learning_rate": 4.122922322885747e-05,
+      "loss": 1.209,
+      "step": 7273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.610051963844261,
+      "learning_rate": 4.122605069733982e-05,
+      "loss": 0.7295,
+      "step": 7274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7252397799207952,
+      "learning_rate": 4.122287771425778e-05,
+      "loss": 1.4824,
+      "step": 7275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9366434349962613,
+      "learning_rate": 4.121970427969967e-05,
+      "loss": 0.8008,
+      "step": 7276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.067691947848654,
+      "learning_rate": 4.1216530393753785e-05,
+      "loss": 1.3242,
+      "step": 7277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1086500264219863,
+      "learning_rate": 4.1213356056508455e-05,
+      "loss": 1.0244,
+      "step": 7278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.728804314317874,
+      "learning_rate": 4.1210181268052035e-05,
+      "loss": 1.2549,
+      "step": 7279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3146813903785644,
+      "learning_rate": 4.120700602847286e-05,
+      "loss": 0.998,
+      "step": 7280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.716858304369315,
+      "learning_rate": 4.1203830337859314e-05,
+      "loss": 1.6074,
+      "step": 7281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.383803746807941,
+      "learning_rate": 4.120065419629976e-05,
+      "loss": 1.3496,
+      "step": 7282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.635416887808528,
+      "learning_rate": 4.11974776038826e-05,
+      "loss": 0.8678,
+      "step": 7283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3852846278702522,
+      "learning_rate": 4.119430056069623e-05,
+      "loss": 1.0803,
+      "step": 7284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0647100687503968,
+      "learning_rate": 4.1191123066829066e-05,
+      "loss": 0.7,
+      "step": 7285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.583387517873554,
+      "learning_rate": 4.1187945122369544e-05,
+      "loss": 1.5254,
+      "step": 7286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.942945586686286,
+      "learning_rate": 4.11847667274061e-05,
+      "loss": 1.0552,
+      "step": 7287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.218811034577575,
+      "learning_rate": 4.118158788202718e-05,
+      "loss": 0.7815,
+      "step": 7288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.203234081408616,
+      "learning_rate": 4.117840858632126e-05,
+      "loss": 1.0264,
+      "step": 7289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1048986338947935,
+      "learning_rate": 4.1175228840376815e-05,
+      "loss": 0.5938,
+      "step": 7290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4965223920395903,
+      "learning_rate": 4.117204864428233e-05,
+      "loss": 1.1914,
+      "step": 7291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7696968750061703,
+      "learning_rate": 4.1168867998126326e-05,
+      "loss": 1.4141,
+      "step": 7292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.662595782617231,
+      "learning_rate": 4.116568690199731e-05,
+      "loss": 1.5,
+      "step": 7293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.624210271642446,
+      "learning_rate": 4.1162505355983795e-05,
+      "loss": 1.0176,
+      "step": 7294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.153071285041086,
+      "learning_rate": 4.115932336017434e-05,
+      "loss": 0.9062,
+      "step": 7295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.117117207312573,
+      "learning_rate": 4.11561409146575e-05,
+      "loss": 1.5918,
+      "step": 7296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9132406533398152,
+      "learning_rate": 4.1152958019521835e-05,
+      "loss": 1.5117,
+      "step": 7297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8871400231023454,
+      "learning_rate": 4.114977467485592e-05,
+      "loss": 1.0059,
+      "step": 7298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6195057181397683,
+      "learning_rate": 4.114659088074835e-05,
+      "loss": 1.0278,
+      "step": 7299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.136286236537119,
+      "learning_rate": 4.1143406637287735e-05,
+      "loss": 1.3442,
+      "step": 7300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6331096900261244,
+      "learning_rate": 4.114022194456268e-05,
+      "loss": 1.1025,
+      "step": 7301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.667904812053224,
+      "learning_rate": 4.113703680266182e-05,
+      "loss": 1.0723,
+      "step": 7302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.052188563350095,
+      "learning_rate": 4.113385121167379e-05,
+      "loss": 1.0132,
+      "step": 7303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4782356843607047,
+      "learning_rate": 4.113066517168726e-05,
+      "loss": 1.165,
+      "step": 7304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7340159688960055,
+      "learning_rate": 4.1127478682790876e-05,
+      "loss": 0.7925,
+      "step": 7305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9872472844416857,
+      "learning_rate": 4.112429174507333e-05,
+      "loss": 1.3083,
+      "step": 7306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.244736979822256,
+      "learning_rate": 4.1121104358623314e-05,
+      "loss": 1.2178,
+      "step": 7307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.347104101773418,
+      "learning_rate": 4.111791652352952e-05,
+      "loss": 1.5684,
+      "step": 7308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.875706158222881,
+      "learning_rate": 4.111472823988067e-05,
+      "loss": 0.8984,
+      "step": 7309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.413683914077992,
+      "learning_rate": 4.1111539507765495e-05,
+      "loss": 0.957,
+      "step": 7310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.481194803265896,
+      "learning_rate": 4.110835032727274e-05,
+      "loss": 0.9863,
+      "step": 7311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9231553813628446,
+      "learning_rate": 4.110516069849114e-05,
+      "loss": 0.9609,
+      "step": 7312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.803023245675265,
+      "learning_rate": 4.1101970621509475e-05,
+      "loss": 1.001,
+      "step": 7313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2257921532344795,
+      "learning_rate": 4.109878009641653e-05,
+      "loss": 0.8379,
+      "step": 7314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.135398012722937,
+      "learning_rate": 4.109558912330109e-05,
+      "loss": 0.8379,
+      "step": 7315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.786526250754257,
+      "learning_rate": 4.109239770225196e-05,
+      "loss": 1.4229,
+      "step": 7316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2942097049901937,
+      "learning_rate": 4.1089205833357946e-05,
+      "loss": 1.2207,
+      "step": 7317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2617988622272454,
+      "learning_rate": 4.108601351670789e-05,
+      "loss": 1.0391,
+      "step": 7318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.167226303402275,
+      "learning_rate": 4.108282075239062e-05,
+      "loss": 0.9805,
+      "step": 7319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.183843860348253,
+      "learning_rate": 4.1079627540495e-05,
+      "loss": 1.8828,
+      "step": 7320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.396779275167687,
+      "learning_rate": 4.107643388110989e-05,
+      "loss": 1.5996,
+      "step": 7321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.184396776955781,
+      "learning_rate": 4.1073239774324166e-05,
+      "loss": 0.793,
+      "step": 7322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.131255983472407,
+      "learning_rate": 4.1070045220226724e-05,
+      "loss": 0.7266,
+      "step": 7323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.115866498813666,
+      "learning_rate": 4.106685021890647e-05,
+      "loss": 1.0483,
+      "step": 7324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.181796187783061,
+      "learning_rate": 4.106365477045232e-05,
+      "loss": 1.1416,
+      "step": 7325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.97686760076144,
+      "learning_rate": 4.106045887495318e-05,
+      "loss": 1.4512,
+      "step": 7326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.310895567228342,
+      "learning_rate": 4.105726253249802e-05,
+      "loss": 1.3379,
+      "step": 7327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.599617629177824,
+      "learning_rate": 4.1054065743175786e-05,
+      "loss": 0.8901,
+      "step": 7328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3969834483454084,
+      "learning_rate": 4.105086850707543e-05,
+      "loss": 0.6628,
+      "step": 7329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3132011103455845,
+      "learning_rate": 4.104767082428594e-05,
+      "loss": 0.5903,
+      "step": 7330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4985054317016067,
+      "learning_rate": 4.1044472694896296e-05,
+      "loss": 1.1309,
+      "step": 7331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.05421847826217,
+      "learning_rate": 4.104127411899552e-05,
+      "loss": 1.2891,
+      "step": 7332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.332223294124987,
+      "learning_rate": 4.103807509667261e-05,
+      "loss": 1.0742,
+      "step": 7333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0058186052407945,
+      "learning_rate": 4.10348756280166e-05,
+      "loss": 0.8398,
+      "step": 7334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7352705006843587,
+      "learning_rate": 4.1031675713116524e-05,
+      "loss": 0.9399,
+      "step": 7335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4876815092372597,
+      "learning_rate": 4.1028475352061445e-05,
+      "loss": 1.0088,
+      "step": 7336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8247211690036216,
+      "learning_rate": 4.1025274544940426e-05,
+      "loss": 1.332,
+      "step": 7337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.230207705211833,
+      "learning_rate": 4.102207329184253e-05,
+      "loss": 0.8613,
+      "step": 7338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.903455193356207,
+      "learning_rate": 4.101887159285687e-05,
+      "loss": 1.6797,
+      "step": 7339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0255787246159302,
+      "learning_rate": 4.1015669448072524e-05,
+      "loss": 0.4944,
+      "step": 7340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8197359801144226,
+      "learning_rate": 4.101246685757862e-05,
+      "loss": 0.9541,
+      "step": 7341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.472617215242659,
+      "learning_rate": 4.100926382146428e-05,
+      "loss": 0.9731,
+      "step": 7342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.177810618314334,
+      "learning_rate": 4.100606033981865e-05,
+      "loss": 1.0537,
+      "step": 7343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.727810248928204,
+      "learning_rate": 4.100285641273088e-05,
+      "loss": 1.0459,
+      "step": 7344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.268925217551159,
+      "learning_rate": 4.099965204029012e-05,
+      "loss": 1.4648,
+      "step": 7345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.022975265468614,
+      "learning_rate": 4.099644722258557e-05,
+      "loss": 1.3867,
+      "step": 7346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.088756048652282,
+      "learning_rate": 4.09932419597064e-05,
+      "loss": 1.041,
+      "step": 7347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9685685724555935,
+      "learning_rate": 4.099003625174181e-05,
+      "loss": 1.0762,
+      "step": 7348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.180000461220338,
+      "learning_rate": 4.0986830098781024e-05,
+      "loss": 1.4824,
+      "step": 7349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.628921395381475,
+      "learning_rate": 4.0983623500913264e-05,
+      "loss": 1.2578,
+      "step": 7350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0862348823159618,
+      "learning_rate": 4.098041645822778e-05,
+      "loss": 0.6143,
+      "step": 7351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.399657282666505,
+      "learning_rate": 4.0977208970813795e-05,
+      "loss": 1.3906,
+      "step": 7352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.08772848679993,
+      "learning_rate": 4.09740010387606e-05,
+      "loss": 1.0703,
+      "step": 7353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8780509259252676,
+      "learning_rate": 4.097079266215746e-05,
+      "loss": 1.1084,
+      "step": 7354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.72617458594013,
+      "learning_rate": 4.096758384109366e-05,
+      "loss": 1.7051,
+      "step": 7355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8567940158419183,
+      "learning_rate": 4.09643745756585e-05,
+      "loss": 0.8799,
+      "step": 7356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2670250040854194,
+      "learning_rate": 4.0961164865941296e-05,
+      "loss": 0.96,
+      "step": 7357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.478734810042449,
+      "learning_rate": 4.095795471203136e-05,
+      "loss": 1.0156,
+      "step": 7358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9060236061611757,
+      "learning_rate": 4.095474411401806e-05,
+      "loss": 0.707,
+      "step": 7359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.268445594059511,
+      "learning_rate": 4.095153307199072e-05,
+      "loss": 1.1133,
+      "step": 7360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5899083746532656,
+      "learning_rate": 4.0948321586038704e-05,
+      "loss": 1.1436,
+      "step": 7361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.293166299917463,
+      "learning_rate": 4.094510965625139e-05,
+      "loss": 1.1934,
+      "step": 7362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.208132798858899,
+      "learning_rate": 4.094189728271817e-05,
+      "loss": 1.1914,
+      "step": 7363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.982029841772485,
+      "learning_rate": 4.093868446552844e-05,
+      "loss": 0.8457,
+      "step": 7364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5551614584176012,
+      "learning_rate": 4.09354712047716e-05,
+      "loss": 1.0024,
+      "step": 7365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.027588415523697,
+      "learning_rate": 4.0932257500537095e-05,
+      "loss": 1.4766,
+      "step": 7366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7903782033691213,
+      "learning_rate": 4.092904335291435e-05,
+      "loss": 0.9902,
+      "step": 7367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.403356806201347,
+      "learning_rate": 4.092582876199281e-05,
+      "loss": 0.7588,
+      "step": 7368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.392832797543941,
+      "learning_rate": 4.092261372786193e-05,
+      "loss": 1.2441,
+      "step": 7369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.750893564867428,
+      "learning_rate": 4.0919398250611206e-05,
+      "loss": 0.8418,
+      "step": 7370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3675089459204846,
+      "learning_rate": 4.09161823303301e-05,
+      "loss": 0.9336,
+      "step": 7371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.605427480024768,
+      "learning_rate": 4.0912965967108125e-05,
+      "loss": 0.991,
+      "step": 7372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.809014044142006,
+      "learning_rate": 4.090974916103478e-05,
+      "loss": 1.5049,
+      "step": 7373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.306380753091258,
+      "learning_rate": 4.09065319121996e-05,
+      "loss": 1.6465,
+      "step": 7374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0249219479059146,
+      "learning_rate": 4.090331422069211e-05,
+      "loss": 0.7051,
+      "step": 7375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3138612073658984,
+      "learning_rate": 4.0900096086601855e-05,
+      "loss": 1.0266,
+      "step": 7376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.193811093555205,
+      "learning_rate": 4.08968775100184e-05,
+      "loss": 1.625,
+      "step": 7377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9745997759758338,
+      "learning_rate": 4.089365849103132e-05,
+      "loss": 0.7148,
+      "step": 7378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9396224774719957,
+      "learning_rate": 4.089043902973019e-05,
+      "loss": 0.8369,
+      "step": 7379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1372812705755933,
+      "learning_rate": 4.088721912620461e-05,
+      "loss": 0.902,
+      "step": 7380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6472033660668983,
+      "learning_rate": 4.08839987805442e-05,
+      "loss": 1.082,
+      "step": 7381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3836928445478116,
+      "learning_rate": 4.0880777992838556e-05,
+      "loss": 1.0869,
+      "step": 7382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.216648072461231,
+      "learning_rate": 4.0877556763177325e-05,
+      "loss": 1.124,
+      "step": 7383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.754041658853676,
+      "learning_rate": 4.087433509165016e-05,
+      "loss": 1.1348,
+      "step": 7384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3332358975729703,
+      "learning_rate": 4.087111297834671e-05,
+      "loss": 1.0,
+      "step": 7385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.210611049302065,
+      "learning_rate": 4.086789042335666e-05,
+      "loss": 1.1372,
+      "step": 7386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.100920625250344,
+      "learning_rate": 4.086466742676967e-05,
+      "loss": 0.4495,
+      "step": 7387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.716162228313839,
+      "learning_rate": 4.0861443988675436e-05,
+      "loss": 0.876,
+      "step": 7388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0010768229367217,
+      "learning_rate": 4.085822010916368e-05,
+      "loss": 1.0254,
+      "step": 7389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9591582840069899,
+      "learning_rate": 4.085499578832411e-05,
+      "loss": 0.5681,
+      "step": 7390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0125401273719308,
+      "learning_rate": 4.085177102624647e-05,
+      "loss": 1.1631,
+      "step": 7391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.277537649391719,
+      "learning_rate": 4.084854582302048e-05,
+      "loss": 1.0664,
+      "step": 7392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.757127022856585,
+      "learning_rate": 4.084532017873593e-05,
+      "loss": 1.3828,
+      "step": 7393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.091536529626275,
+      "learning_rate": 4.0842094093482564e-05,
+      "loss": 1.2861,
+      "step": 7394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6767899665446104,
+      "learning_rate": 4.0838867567350165e-05,
+      "loss": 1.0166,
+      "step": 7395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5031979134711557,
+      "learning_rate": 4.083564060042853e-05,
+      "loss": 1.4395,
+      "step": 7396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0792253551735795,
+      "learning_rate": 4.083241319280746e-05,
+      "loss": 0.8828,
+      "step": 7397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4020170455676024,
+      "learning_rate": 4.082918534457679e-05,
+      "loss": 1.2383,
+      "step": 7398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1930508621656295,
+      "learning_rate": 4.082595705582632e-05,
+      "loss": 0.941,
+      "step": 7399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.387885007168857,
+      "learning_rate": 4.082272832664592e-05,
+      "loss": 1.7188,
+      "step": 7400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0290771247261206,
+      "learning_rate": 4.0819499157125427e-05,
+      "loss": 1.6582,
+      "step": 7401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0078115500362537,
+      "learning_rate": 4.0816269547354716e-05,
+      "loss": 1.1504,
+      "step": 7402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.613444643942531,
+      "learning_rate": 4.081303949742365e-05,
+      "loss": 1.2285,
+      "step": 7403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.149702383396289,
+      "learning_rate": 4.080980900742215e-05,
+      "loss": 1.877,
+      "step": 7404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.055618963643301,
+      "learning_rate": 4.0806578077440085e-05,
+      "loss": 1.1709,
+      "step": 7405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.306610073637671,
+      "learning_rate": 4.08033467075674e-05,
+      "loss": 0.3921,
+      "step": 7406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.816983599064081,
+      "learning_rate": 4.0800114897894004e-05,
+      "loss": 1.1387,
+      "step": 7407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.083391950092155,
+      "learning_rate": 4.079688264850985e-05,
+      "loss": 1.022,
+      "step": 7408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8122967525149245,
+      "learning_rate": 4.079364995950487e-05,
+      "loss": 1.666,
+      "step": 7409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.459513127908648,
+      "learning_rate": 4.079041683096905e-05,
+      "loss": 1.0155,
+      "step": 7410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.911723588540941,
+      "learning_rate": 4.0787183262992354e-05,
+      "loss": 0.5386,
+      "step": 7411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.029588562387226,
+      "learning_rate": 4.0783949255664774e-05,
+      "loss": 1.793,
+      "step": 7412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.567073731529146,
+      "learning_rate": 4.0780714809076316e-05,
+      "loss": 1.7861,
+      "step": 7413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.268652320629627,
+      "learning_rate": 4.077747992331699e-05,
+      "loss": 1.0156,
+      "step": 7414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2688540323598865,
+      "learning_rate": 4.077424459847681e-05,
+      "loss": 1.2373,
+      "step": 7415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0038860900758,
+      "learning_rate": 4.077100883464584e-05,
+      "loss": 1.3652,
+      "step": 7416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7514856833588577,
+      "learning_rate": 4.07677726319141e-05,
+      "loss": 1.0654,
+      "step": 7417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.602027828539811,
+      "learning_rate": 4.076453599037168e-05,
+      "loss": 1.418,
+      "step": 7418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.315543477476863,
+      "learning_rate": 4.0761298910108625e-05,
+      "loss": 1.209,
+      "step": 7419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.199108039900857,
+      "learning_rate": 4.0758061391215054e-05,
+      "loss": 1.3154,
+      "step": 7420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9699899311098816,
+      "learning_rate": 4.0754823433781044e-05,
+      "loss": 1.0566,
+      "step": 7421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.707189161750841,
+      "learning_rate": 4.075158503789671e-05,
+      "loss": 1.0742,
+      "step": 7422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.916301886196543,
+      "learning_rate": 4.0748346203652176e-05,
+      "loss": 1.2367,
+      "step": 7423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.513623155104489,
+      "learning_rate": 4.074510693113758e-05,
+      "loss": 1.2832,
+      "step": 7424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.279550925252108,
+      "learning_rate": 4.074186722044306e-05,
+      "loss": 1.0713,
+      "step": 7425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.677694570657208,
+      "learning_rate": 4.0738627071658795e-05,
+      "loss": 1.291,
+      "step": 7426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3576089678891177,
+      "learning_rate": 4.073538648487495e-05,
+      "loss": 0.5631,
+      "step": 7427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.399185035832454,
+      "learning_rate": 4.073214546018169e-05,
+      "loss": 1.001,
+      "step": 7428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6678847471447957,
+      "learning_rate": 4.072890399766923e-05,
+      "loss": 1.1816,
+      "step": 7429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9179517197484113,
+      "learning_rate": 4.072566209742779e-05,
+      "loss": 0.9004,
+      "step": 7430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9660452316858303,
+      "learning_rate": 4.0722419759547556e-05,
+      "loss": 1.0566,
+      "step": 7431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.873360578266876,
+      "learning_rate": 4.071917698411879e-05,
+      "loss": 1.0752,
+      "step": 7432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.899118677589879,
+      "learning_rate": 4.071593377123173e-05,
+      "loss": 1.0713,
+      "step": 7433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.391870939557701,
+      "learning_rate": 4.071269012097663e-05,
+      "loss": 1.1992,
+      "step": 7434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7442367156774874,
+      "learning_rate": 4.0709446033443754e-05,
+      "loss": 1.3711,
+      "step": 7435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.144394045472276,
+      "learning_rate": 4.0706201508723397e-05,
+      "loss": 1.2715,
+      "step": 7436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.10413608591752,
+      "learning_rate": 4.070295654690584e-05,
+      "loss": 1.0674,
+      "step": 7437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.557441600285368,
+      "learning_rate": 4.06997111480814e-05,
+      "loss": 0.5974,
+      "step": 7438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6164673682768935,
+      "learning_rate": 4.0696465312340394e-05,
+      "loss": 1.3203,
+      "step": 7439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.454726571184091,
+      "learning_rate": 4.069321903977314e-05,
+      "loss": 1.4995,
+      "step": 7440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.630177319401945,
+      "learning_rate": 4.0689972330469994e-05,
+      "loss": 0.8955,
+      "step": 7441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.756777424526457,
+      "learning_rate": 4.06867251845213e-05,
+      "loss": 0.7754,
+      "step": 7442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1032923719555336,
+      "learning_rate": 4.068347760201744e-05,
+      "loss": 0.9072,
+      "step": 7443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.271769642649786,
+      "learning_rate": 4.0680229583048776e-05,
+      "loss": 0.6841,
+      "step": 7444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2258891218113344,
+      "learning_rate": 4.0676981127705706e-05,
+      "loss": 1.5293,
+      "step": 7445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8889836780777607,
+      "learning_rate": 4.067373223607863e-05,
+      "loss": 1.3043,
+      "step": 7446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4837508467205813,
+      "learning_rate": 4.0670482908257966e-05,
+      "loss": 1.0703,
+      "step": 7447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.527290827154309,
+      "learning_rate": 4.066723314433415e-05,
+      "loss": 1.3945,
+      "step": 7448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.124733743913925,
+      "learning_rate": 4.0663982944397616e-05,
+      "loss": 1.8145,
+      "step": 7449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7434544258538467,
+      "learning_rate": 4.066073230853881e-05,
+      "loss": 0.7566,
+      "step": 7450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7079978637292457,
+      "learning_rate": 4.0657481236848196e-05,
+      "loss": 1.2246,
+      "step": 7451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8209234039352857,
+      "learning_rate": 4.065422972941625e-05,
+      "loss": 1.0569,
+      "step": 7452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9058372175707792,
+      "learning_rate": 4.065097778633347e-05,
+      "loss": 0.731,
+      "step": 7453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.864621445844047,
+      "learning_rate": 4.064772540769035e-05,
+      "loss": 1.4492,
+      "step": 7454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2378244516420827,
+      "learning_rate": 4.0644472593577396e-05,
+      "loss": 1.0215,
+      "step": 7455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0134150654615253,
+      "learning_rate": 4.0641219344085144e-05,
+      "loss": 1.3691,
+      "step": 7456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.350526493857101,
+      "learning_rate": 4.0637965659304124e-05,
+      "loss": 2.2383,
+      "step": 7457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8687643345825826,
+      "learning_rate": 4.063471153932489e-05,
+      "loss": 1.4443,
+      "step": 7458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5641019512444765,
+      "learning_rate": 4.063145698423799e-05,
+      "loss": 1.541,
+      "step": 7459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.940655737948327,
+      "learning_rate": 4.0628201994134016e-05,
+      "loss": 0.8682,
+      "step": 7460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6689997143021706,
+      "learning_rate": 4.062494656910353e-05,
+      "loss": 1.6504,
+      "step": 7461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.850306884128497,
+      "learning_rate": 4.062169070923716e-05,
+      "loss": 1.2046,
+      "step": 7462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.586770081204423,
+      "learning_rate": 4.0618434414625484e-05,
+      "loss": 1.043,
+      "step": 7463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.948543522212272,
+      "learning_rate": 4.061517768535913e-05,
+      "loss": 1.501,
+      "step": 7464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.564765858810987,
+      "learning_rate": 4.061192052152875e-05,
+      "loss": 1.1582,
+      "step": 7465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5132897952868016,
+      "learning_rate": 4.060866292322497e-05,
+      "loss": 1.3594,
+      "step": 7466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1895394763371407,
+      "learning_rate": 4.060540489053847e-05,
+      "loss": 0.8063,
+      "step": 7467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3455085764638044,
+      "learning_rate": 4.060214642355989e-05,
+      "loss": 0.7725,
+      "step": 7468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.687814602120315,
+      "learning_rate": 4.059888752237993e-05,
+      "loss": 0.8252,
+      "step": 7469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2302368898685225,
+      "learning_rate": 4.059562818708928e-05,
+      "loss": 0.9082,
+      "step": 7470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7732593747816763,
+      "learning_rate": 4.0592368417778646e-05,
+      "loss": 1.2773,
+      "step": 7471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.100444396999472,
+      "learning_rate": 4.058910821453875e-05,
+      "loss": 1.5156,
+      "step": 7472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.090238786955433,
+      "learning_rate": 4.0585847577460316e-05,
+      "loss": 1.1113,
+      "step": 7473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8070192672259624,
+      "learning_rate": 4.058258650663409e-05,
+      "loss": 0.8893,
+      "step": 7474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.309879150164857,
+      "learning_rate": 4.057932500215083e-05,
+      "loss": 1.1631,
+      "step": 7475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.894760275299884,
+      "learning_rate": 4.057606306410129e-05,
+      "loss": 0.5464,
+      "step": 7476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.051352126856679,
+      "learning_rate": 4.057280069257626e-05,
+      "loss": 0.875,
+      "step": 7477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6840419256230326,
+      "learning_rate": 4.056953788766653e-05,
+      "loss": 1.6211,
+      "step": 7478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.000040212997494,
+      "learning_rate": 4.0566274649462885e-05,
+      "loss": 1.416,
+      "step": 7479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.8088915306188,
+      "learning_rate": 4.056301097805616e-05,
+      "loss": 1.2012,
+      "step": 7480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.026298709309806,
+      "learning_rate": 4.055974687353719e-05,
+      "loss": 0.8613,
+      "step": 7481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8079043346377452,
+      "learning_rate": 4.055648233599678e-05,
+      "loss": 1.1826,
+      "step": 7482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7878789193335056,
+      "learning_rate": 4.055321736552581e-05,
+      "loss": 0.8345,
+      "step": 7483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.273915042730783,
+      "learning_rate": 4.0549951962215126e-05,
+      "loss": 1.043,
+      "step": 7484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7524280233093354,
+      "learning_rate": 4.054668612615561e-05,
+      "loss": 1.0273,
+      "step": 7485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.772636883362757,
+      "learning_rate": 4.054341985743815e-05,
+      "loss": 0.8945,
+      "step": 7486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9826377086072093,
+      "learning_rate": 4.054015315615365e-05,
+      "loss": 0.8496,
+      "step": 7487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5007141047064505,
+      "learning_rate": 4.0536886022393006e-05,
+      "loss": 0.6056,
+      "step": 7488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.028623912805836,
+      "learning_rate": 4.0533618456247145e-05,
+      "loss": 1.3672,
+      "step": 7489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0881028955106573,
+      "learning_rate": 4.053035045780701e-05,
+      "loss": 0.8975,
+      "step": 7490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1971847858393203,
+      "learning_rate": 4.052708202716355e-05,
+      "loss": 0.9199,
+      "step": 7491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.804904174935841,
+      "learning_rate": 4.0523813164407715e-05,
+      "loss": 0.9849,
+      "step": 7492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.001789240617995,
+      "learning_rate": 4.0520543869630475e-05,
+      "loss": 1.8496,
+      "step": 7493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6753713474812852,
+      "learning_rate": 4.0517274142922825e-05,
+      "loss": 1.0312,
+      "step": 7494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.354970901536263,
+      "learning_rate": 4.051400398437575e-05,
+      "loss": 1.209,
+      "step": 7495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9234942858622652,
+      "learning_rate": 4.051073339408026e-05,
+      "loss": 1.416,
+      "step": 7496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.492404844558702,
+      "learning_rate": 4.050746237212737e-05,
+      "loss": 1.2041,
+      "step": 7497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7953670317186115,
+      "learning_rate": 4.0504190918608116e-05,
+      "loss": 0.5222,
+      "step": 7498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7807484443881814,
+      "learning_rate": 4.0500919033613536e-05,
+      "loss": 0.7939,
+      "step": 7499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.216167062026457,
+      "learning_rate": 4.0497646717234705e-05,
+      "loss": 1.5801,
+      "step": 7500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.596876902298397,
+      "learning_rate": 4.049437396956266e-05,
+      "loss": 1.0664,
+      "step": 7501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1666408561734203,
+      "learning_rate": 4.049110079068851e-05,
+      "loss": 0.8252,
+      "step": 7502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.666368398277562,
+      "learning_rate": 4.048782718070332e-05,
+      "loss": 1.2617,
+      "step": 7503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0794933395301327,
+      "learning_rate": 4.04845531396982e-05,
+      "loss": 1.6914,
+      "step": 7504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6232486513745656,
+      "learning_rate": 4.048127866776428e-05,
+      "loss": 0.769,
+      "step": 7505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.821226061246514,
+      "learning_rate": 4.047800376499268e-05,
+      "loss": 1.0469,
+      "step": 7506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.672605894183615,
+      "learning_rate": 4.047472843147452e-05,
+      "loss": 0.9541,
+      "step": 7507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.8208860525409785,
+      "learning_rate": 4.047145266730098e-05,
+      "loss": 1.2266,
+      "step": 7508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1407561772978134,
+      "learning_rate": 4.046817647256321e-05,
+      "loss": 1.1079,
+      "step": 7509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.451545515526786,
+      "learning_rate": 4.046489984735239e-05,
+      "loss": 0.9023,
+      "step": 7510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.0300705241429,
+      "learning_rate": 4.0461622791759696e-05,
+      "loss": 1.6992,
+      "step": 7511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7570270819907607,
+      "learning_rate": 4.045834530587633e-05,
+      "loss": 1.21,
+      "step": 7512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.942096509877514,
+      "learning_rate": 4.045506738979352e-05,
+      "loss": 1.1309,
+      "step": 7513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.849756466758432,
+      "learning_rate": 4.0451789043602465e-05,
+      "loss": 0.803,
+      "step": 7514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1636498674112636,
+      "learning_rate": 4.044851026739442e-05,
+      "loss": 0.9285,
+      "step": 7515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9795987176294003,
+      "learning_rate": 4.0445231061260616e-05,
+      "loss": 1.0781,
+      "step": 7516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.380098859430143,
+      "learning_rate": 4.0441951425292325e-05,
+      "loss": 1.6816,
+      "step": 7517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4197838399834133,
+      "learning_rate": 4.0438671359580804e-05,
+      "loss": 1.1426,
+      "step": 7518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.7102884380035,
+      "learning_rate": 4.043539086421736e-05,
+      "loss": 0.9712,
+      "step": 7519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.98159377612155,
+      "learning_rate": 4.043210993929326e-05,
+      "loss": 1.3926,
+      "step": 7520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.782618108747665,
+      "learning_rate": 4.0428828584899816e-05,
+      "loss": 1.2871,
+      "step": 7521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.746653775104651,
+      "learning_rate": 4.042554680112837e-05,
+      "loss": 0.8145,
+      "step": 7522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.632669162998846,
+      "learning_rate": 4.0422264588070224e-05,
+      "loss": 1.3418,
+      "step": 7523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1570143519716822,
+      "learning_rate": 4.041898194581674e-05,
+      "loss": 1.0273,
+      "step": 7524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6910406343247883,
+      "learning_rate": 4.0415698874459265e-05,
+      "loss": 0.9834,
+      "step": 7525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.152157756433166,
+      "learning_rate": 4.041241537408916e-05,
+      "loss": 1.2598,
+      "step": 7526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8638261013834,
+      "learning_rate": 4.040913144479781e-05,
+      "loss": 0.6626,
+      "step": 7527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.067048797997507,
+      "learning_rate": 4.0405847086676604e-05,
+      "loss": 1.4229,
+      "step": 7528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.21987198950306,
+      "learning_rate": 4.040256229981696e-05,
+      "loss": 0.7507,
+      "step": 7529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9450516724384515,
+      "learning_rate": 4.039927708431026e-05,
+      "loss": 1.8105,
+      "step": 7530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.496564268704447,
+      "learning_rate": 4.039599144024796e-05,
+      "loss": 0.877,
+      "step": 7531
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.839042875970548,
+      "learning_rate": 4.039270536772148e-05,
+      "loss": 1.0518,
+      "step": 7532
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.590846590909421,
+      "learning_rate": 4.038941886682228e-05,
+      "loss": 0.8135,
+      "step": 7533
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6848820590603126,
+      "learning_rate": 4.038613193764181e-05,
+      "loss": 1.1885,
+      "step": 7534
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.16723070383497,
+      "learning_rate": 4.0382844580271565e-05,
+      "loss": 0.8368,
+      "step": 7535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.698755222666605,
+      "learning_rate": 4.0379556794803e-05,
+      "loss": 1.5664,
+      "step": 7536
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.844202529617042,
+      "learning_rate": 4.037626858132765e-05,
+      "loss": 0.959,
+      "step": 7537
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8573537622546596,
+      "learning_rate": 4.037297993993699e-05,
+      "loss": 0.9648,
+      "step": 7538
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.414343448191626,
+      "learning_rate": 4.036969087072257e-05,
+      "loss": 0.6025,
+      "step": 7539
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2132641307897214,
+      "learning_rate": 4.036640137377588e-05,
+      "loss": 1.5566,
+      "step": 7540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.899353409010501,
+      "learning_rate": 4.036311144918853e-05,
+      "loss": 1.3623,
+      "step": 7541
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1954839609119934,
+      "learning_rate": 4.035982109705203e-05,
+      "loss": 1.1094,
+      "step": 7542
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3750172664617013,
+      "learning_rate": 4.035653031745796e-05,
+      "loss": 1.1348,
+      "step": 7543
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.882382672264412,
+      "learning_rate": 4.0353239110497906e-05,
+      "loss": 0.6016,
+      "step": 7544
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.399493314350173,
+      "learning_rate": 4.034994747626346e-05,
+      "loss": 0.6797,
+      "step": 7545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.159954989380313,
+      "learning_rate": 4.034665541484621e-05,
+      "loss": 1.5293,
+      "step": 7546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3117613058737367,
+      "learning_rate": 4.03433629263378e-05,
+      "loss": 0.8604,
+      "step": 7547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0283931889893005,
+      "learning_rate": 4.034007001082985e-05,
+      "loss": 1.4424,
+      "step": 7548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.402349165715946,
+      "learning_rate": 4.033677666841399e-05,
+      "loss": 0.9873,
+      "step": 7549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.01681447659309,
+      "learning_rate": 4.033348289918188e-05,
+      "loss": 1.1406,
+      "step": 7550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.714636215787682,
+      "learning_rate": 4.033018870322519e-05,
+      "loss": 0.9785,
+      "step": 7551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.426675479766163,
+      "learning_rate": 4.032689408063558e-05,
+      "loss": 1.2217,
+      "step": 7552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.440546005066807,
+      "learning_rate": 4.032359903150475e-05,
+      "loss": 2.0566,
+      "step": 7553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.704215576802714,
+      "learning_rate": 4.032030355592439e-05,
+      "loss": 1.541,
+      "step": 7554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.370362576036513,
+      "learning_rate": 4.031700765398623e-05,
+      "loss": 1.2715,
+      "step": 7555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.879202880484414,
+      "learning_rate": 4.031371132578198e-05,
+      "loss": 0.7773,
+      "step": 7556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9746130619539692,
+      "learning_rate": 4.0310414571403376e-05,
+      "loss": 0.6807,
+      "step": 7557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1758136817993443,
+      "learning_rate": 4.030711739094216e-05,
+      "loss": 1.7637,
+      "step": 7558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3273953312039537,
+      "learning_rate": 4.030381978449011e-05,
+      "loss": 1.8496,
+      "step": 7559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.926043632364524,
+      "learning_rate": 4.030052175213899e-05,
+      "loss": 0.8379,
+      "step": 7560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4031450129396172,
+      "learning_rate": 4.0297223293980566e-05,
+      "loss": 0.9741,
+      "step": 7561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.0688363764863,
+      "learning_rate": 4.029392441010665e-05,
+      "loss": 1.3652,
+      "step": 7562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3342593489957415,
+      "learning_rate": 4.029062510060904e-05,
+      "loss": 1.1416,
+      "step": 7563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.564105900051419,
+      "learning_rate": 4.028732536557957e-05,
+      "loss": 1.0269,
+      "step": 7564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.993094603154808,
+      "learning_rate": 4.028402520511004e-05,
+      "loss": 1.2334,
+      "step": 7565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.884736367647421,
+      "learning_rate": 4.0280724619292325e-05,
+      "loss": 1.085,
+      "step": 7566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9418531587469476,
+      "learning_rate": 4.027742360821825e-05,
+      "loss": 1.1553,
+      "step": 7567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.739015485658099,
+      "learning_rate": 4.0274122171979714e-05,
+      "loss": 1.2988,
+      "step": 7568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9575076931982878,
+      "learning_rate": 4.027082031066857e-05,
+      "loss": 0.9785,
+      "step": 7569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.913862539965024,
+      "learning_rate": 4.026751802437672e-05,
+      "loss": 1.4619,
+      "step": 7570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.322409558788553,
+      "learning_rate": 4.0264215313196044e-05,
+      "loss": 1.1797,
+      "step": 7571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5671320662547266,
+      "learning_rate": 4.0260912177218475e-05,
+      "loss": 1.0908,
+      "step": 7572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1645296880140474,
+      "learning_rate": 4.025760861653594e-05,
+      "loss": 0.8182,
+      "step": 7573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.314361956768486,
+      "learning_rate": 4.0254304631240356e-05,
+      "loss": 0.8643,
+      "step": 7574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.522829426691794,
+      "learning_rate": 4.0251000221423694e-05,
+      "loss": 1.1113,
+      "step": 7575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7444072553128693,
+      "learning_rate": 4.0247695387177904e-05,
+      "loss": 0.8115,
+      "step": 7576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7297191246171817,
+      "learning_rate": 4.024439012859495e-05,
+      "loss": 0.8872,
+      "step": 7577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.231905682124859,
+      "learning_rate": 4.024108444576683e-05,
+      "loss": 0.7017,
+      "step": 7578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.240631094994761,
+      "learning_rate": 4.0237778338785534e-05,
+      "loss": 1.0747,
+      "step": 7579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.103655398966165,
+      "learning_rate": 4.0234471807743076e-05,
+      "loss": 1.6094,
+      "step": 7580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9193412391726792,
+      "learning_rate": 4.023116485273146e-05,
+      "loss": 0.8456,
+      "step": 7581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.510656278142714,
+      "learning_rate": 4.022785747384274e-05,
+      "loss": 1.5449,
+      "step": 7582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.00525689664881,
+      "learning_rate": 4.022454967116893e-05,
+      "loss": 1.0127,
+      "step": 7583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4527913618515327,
+      "learning_rate": 4.022124144480211e-05,
+      "loss": 1.1104,
+      "step": 7584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.723306805291962,
+      "learning_rate": 4.021793279483434e-05,
+      "loss": 1.1792,
+      "step": 7585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.94802322176797,
+      "learning_rate": 4.021462372135769e-05,
+      "loss": 0.4868,
+      "step": 7586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.677362762491946,
+      "learning_rate": 4.021131422446426e-05,
+      "loss": 1.4453,
+      "step": 7587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.062301161374918,
+      "learning_rate": 4.020800430424614e-05,
+      "loss": 1.0176,
+      "step": 7588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4024542062068828,
+      "learning_rate": 4.020469396079546e-05,
+      "loss": 0.9473,
+      "step": 7589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.7082323496743985,
+      "learning_rate": 4.020138319420433e-05,
+      "loss": 1.3438,
+      "step": 7590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.239591657785267,
+      "learning_rate": 4.019807200456489e-05,
+      "loss": 1.4961,
+      "step": 7591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.732935761016042,
+      "learning_rate": 4.019476039196931e-05,
+      "loss": 0.8359,
+      "step": 7592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2257014165548568,
+      "learning_rate": 4.019144835650972e-05,
+      "loss": 1.4316,
+      "step": 7593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.636596813476618,
+      "learning_rate": 4.018813589827831e-05,
+      "loss": 1.1885,
+      "step": 7594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.490336164455731,
+      "learning_rate": 4.018482301736726e-05,
+      "loss": 1.5322,
+      "step": 7595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.022091333056549,
+      "learning_rate": 4.018150971386877e-05,
+      "loss": 0.6008,
+      "step": 7596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.272018615496362,
+      "learning_rate": 4.0178195987875045e-05,
+      "loss": 1.4258,
+      "step": 7597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8271111839137997,
+      "learning_rate": 4.017488183947831e-05,
+      "loss": 0.8711,
+      "step": 7598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.966057090312859,
+      "learning_rate": 4.017156726877078e-05,
+      "loss": 1.3647,
+      "step": 7599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3915336726978578,
+      "learning_rate": 4.016825227584471e-05,
+      "loss": 0.7739,
+      "step": 7600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.26957215108756,
+      "learning_rate": 4.016493686079236e-05,
+      "loss": 1.8652,
+      "step": 7601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.956906826165762,
+      "learning_rate": 4.016162102370599e-05,
+      "loss": 1.4375,
+      "step": 7602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.393027751621903,
+      "learning_rate": 4.015830476467788e-05,
+      "loss": 1.3965,
+      "step": 7603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.048903487280773,
+      "learning_rate": 4.0154988083800315e-05,
+      "loss": 1.8311,
+      "step": 7604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.432278346502411,
+      "learning_rate": 4.0151670981165604e-05,
+      "loss": 1.2207,
+      "step": 7605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1555966889175355,
+      "learning_rate": 4.014835345686606e-05,
+      "loss": 1.0918,
+      "step": 7606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.474912653451009,
+      "learning_rate": 4.0145035510994e-05,
+      "loss": 0.8794,
+      "step": 7607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9607256168673657,
+      "learning_rate": 4.014171714364178e-05,
+      "loss": 1.2969,
+      "step": 7608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.388115107790445,
+      "learning_rate": 4.013839835490172e-05,
+      "loss": 1.0771,
+      "step": 7609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.022336820394844,
+      "learning_rate": 4.0135079144866206e-05,
+      "loss": 1.1572,
+      "step": 7610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.293594495242325,
+      "learning_rate": 4.01317595136276e-05,
+      "loss": 0.6816,
+      "step": 7611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.823208576795349,
+      "learning_rate": 4.012843946127828e-05,
+      "loss": 1.0864,
+      "step": 7612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1665425694560376,
+      "learning_rate": 4.0125118987910656e-05,
+      "loss": 1.0566,
+      "step": 7613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.220038778464683,
+      "learning_rate": 4.0121798093617124e-05,
+      "loss": 1.1602,
+      "step": 7614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5470102427237653,
+      "learning_rate": 4.011847677849011e-05,
+      "loss": 1.5742,
+      "step": 7615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.989087052496865,
+      "learning_rate": 4.011515504262204e-05,
+      "loss": 0.7988,
+      "step": 7616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7842426925751134,
+      "learning_rate": 4.011183288610536e-05,
+      "loss": 0.6396,
+      "step": 7617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.587099279037457,
+      "learning_rate": 4.010851030903252e-05,
+      "loss": 1.0868,
+      "step": 7618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.426764861078227,
+      "learning_rate": 4.0105187311495995e-05,
+      "loss": 2.1133,
+      "step": 7619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1460192578208015,
+      "learning_rate": 4.010186389358825e-05,
+      "loss": 0.7484,
+      "step": 7620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.228217029844613,
+      "learning_rate": 4.009854005540179e-05,
+      "loss": 0.54,
+      "step": 7621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.028776570238389,
+      "learning_rate": 4.00952157970291e-05,
+      "loss": 1.0977,
+      "step": 7622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.326926108307995,
+      "learning_rate": 4.0091891118562695e-05,
+      "loss": 1.0913,
+      "step": 7623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.698344772327187,
+      "learning_rate": 4.008856602009511e-05,
+      "loss": 1.3203,
+      "step": 7624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.226296456583726,
+      "learning_rate": 4.008524050171888e-05,
+      "loss": 1.6016,
+      "step": 7625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.48091263988844,
+      "learning_rate": 4.008191456352654e-05,
+      "loss": 0.957,
+      "step": 7626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.247069251480084,
+      "learning_rate": 4.007858820561066e-05,
+      "loss": 1.2207,
+      "step": 7627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6633288136691755,
+      "learning_rate": 4.0075261428063806e-05,
+      "loss": 1.2383,
+      "step": 7628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1713833698429594,
+      "learning_rate": 4.007193423097857e-05,
+      "loss": 1.1274,
+      "step": 7629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.612746424928264,
+      "learning_rate": 4.006860661444753e-05,
+      "loss": 1.2603,
+      "step": 7630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.58461296535472,
+      "learning_rate": 4.006527857856332e-05,
+      "loss": 1.3262,
+      "step": 7631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.45964114297838,
+      "learning_rate": 4.006195012341852e-05,
+      "loss": 1.1172,
+      "step": 7632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7186796683498997,
+      "learning_rate": 4.005862124910579e-05,
+      "loss": 0.7529,
+      "step": 7633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.552065446482524,
+      "learning_rate": 4.005529195571777e-05,
+      "loss": 1.416,
+      "step": 7634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.2667040455690834,
+      "learning_rate": 4.005196224334709e-05,
+      "loss": 1.0967,
+      "step": 7635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7491349812056387,
+      "learning_rate": 4.004863211208643e-05,
+      "loss": 0.7256,
+      "step": 7636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.157118832015152,
+      "learning_rate": 4.004530156202846e-05,
+      "loss": 1.6926,
+      "step": 7637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.198426884975983,
+      "learning_rate": 4.0041970593265886e-05,
+      "loss": 1.0469,
+      "step": 7638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2995334732595465,
+      "learning_rate": 4.003863920589139e-05,
+      "loss": 1.3906,
+      "step": 7639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.407911542863807,
+      "learning_rate": 4.003530739999769e-05,
+      "loss": 0.7178,
+      "step": 7640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2241250499866876,
+      "learning_rate": 4.00319751756775e-05,
+      "loss": 0.8486,
+      "step": 7641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4693756276532732,
+      "learning_rate": 4.0028642533023556e-05,
+      "loss": 1.0273,
+      "step": 7642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9780726638306705,
+      "learning_rate": 4.002530947212861e-05,
+      "loss": 0.6023,
+      "step": 7643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.331561221705384,
+      "learning_rate": 4.0021975993085424e-05,
+      "loss": 1.0439,
+      "step": 7644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5897097225303343,
+      "learning_rate": 4.001864209598676e-05,
+      "loss": 1.1348,
+      "step": 7645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.000265877330339,
+      "learning_rate": 4.0015307780925396e-05,
+      "loss": 1.0576,
+      "step": 7646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.894665178297103,
+      "learning_rate": 4.0011973047994136e-05,
+      "loss": 1.0273,
+      "step": 7647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2997886474436284,
+      "learning_rate": 4.0008637897285775e-05,
+      "loss": 1.2246,
+      "step": 7648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.197023403220178,
+      "learning_rate": 4.000530232889312e-05,
+      "loss": 1.3197,
+      "step": 7649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0464160200755073,
+      "learning_rate": 4.0001966342909025e-05,
+      "loss": 0.7822,
+      "step": 7650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.319623950201519,
+      "learning_rate": 3.999862993942631e-05,
+      "loss": 0.7314,
+      "step": 7651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.384557266568355,
+      "learning_rate": 3.999529311853783e-05,
+      "loss": 0.9092,
+      "step": 7652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7288904160896252,
+      "learning_rate": 3.999195588033645e-05,
+      "loss": 0.6541,
+      "step": 7653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.296629037147203,
+      "learning_rate": 3.998861822491503e-05,
+      "loss": 1.3262,
+      "step": 7654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.47322765181921,
+      "learning_rate": 3.998528015236648e-05,
+      "loss": 1.2441,
+      "step": 7655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.302242210279731,
+      "learning_rate": 3.9981941662783674e-05,
+      "loss": 1.3223,
+      "step": 7656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8599806090347166,
+      "learning_rate": 3.997860275625954e-05,
+      "loss": 0.8369,
+      "step": 7657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.338837088497788,
+      "learning_rate": 3.997526343288698e-05,
+      "loss": 1.2021,
+      "step": 7658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2607726043991625,
+      "learning_rate": 3.997192369275894e-05,
+      "loss": 1.1006,
+      "step": 7659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3639392283611103,
+      "learning_rate": 3.996858353596835e-05,
+      "loss": 0.6963,
+      "step": 7660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.580408922657392,
+      "learning_rate": 3.9965242962608184e-05,
+      "loss": 0.6011,
+      "step": 7661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.3637809018651295,
+      "learning_rate": 3.99619019727714e-05,
+      "loss": 1.458,
+      "step": 7662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8012884683618884,
+      "learning_rate": 3.9958560566550975e-05,
+      "loss": 1.4141,
+      "step": 7663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2557362444412874,
+      "learning_rate": 3.995521874403989e-05,
+      "loss": 0.4907,
+      "step": 7664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.535341699808931,
+      "learning_rate": 3.9951876505331156e-05,
+      "loss": 0.8676,
+      "step": 7665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6458257439772326,
+      "learning_rate": 3.994853385051779e-05,
+      "loss": 0.8457,
+      "step": 7666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.432000020416215,
+      "learning_rate": 3.994519077969282e-05,
+      "loss": 1.5967,
+      "step": 7667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.20146969280409,
+      "learning_rate": 3.9941847292949265e-05,
+      "loss": 0.7734,
+      "step": 7668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.112441466227583,
+      "learning_rate": 3.993850339038019e-05,
+      "loss": 1.6875,
+      "step": 7669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.67276191491399,
+      "learning_rate": 3.9935159072078646e-05,
+      "loss": 0.9629,
+      "step": 7670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.496247050584484,
+      "learning_rate": 3.9931814338137695e-05,
+      "loss": 1.2441,
+      "step": 7671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.559170117870187,
+      "learning_rate": 3.9928469188650444e-05,
+      "loss": 0.7292,
+      "step": 7672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6501306393649133,
+      "learning_rate": 3.992512362370996e-05,
+      "loss": 0.8855,
+      "step": 7673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.580277163085835,
+      "learning_rate": 3.992177764340936e-05,
+      "loss": 1.2188,
+      "step": 7674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3160012005224724,
+      "learning_rate": 3.991843124784178e-05,
+      "loss": 1.0547,
+      "step": 7675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.716339358015499,
+      "learning_rate": 3.991508443710032e-05,
+      "loss": 0.7866,
+      "step": 7676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.686907726931122,
+      "learning_rate": 3.991173721127813e-05,
+      "loss": 0.7383,
+      "step": 7677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.272382700176516,
+      "learning_rate": 3.990838957046837e-05,
+      "loss": 0.7803,
+      "step": 7678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.663526914899603,
+      "learning_rate": 3.990504151476419e-05,
+      "loss": 1.2344,
+      "step": 7679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.994493283327756,
+      "learning_rate": 3.990169304425878e-05,
+      "loss": 1.5039,
+      "step": 7680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.283690120230472,
+      "learning_rate": 3.989834415904532e-05,
+      "loss": 1.4219,
+      "step": 7681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.704569260085746,
+      "learning_rate": 3.9894994859217e-05,
+      "loss": 0.8315,
+      "step": 7682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.2284676539287025,
+      "learning_rate": 3.989164514486704e-05,
+      "loss": 1.3945,
+      "step": 7683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.430118534823402,
+      "learning_rate": 3.9888295016088654e-05,
+      "loss": 0.8516,
+      "step": 7684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.838731596579442,
+      "learning_rate": 3.988494447297508e-05,
+      "loss": 1.2729,
+      "step": 7685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.0853618055752445,
+      "learning_rate": 3.988159351561957e-05,
+      "loss": 1.1855,
+      "step": 7686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.302259504616958,
+      "learning_rate": 3.987824214411537e-05,
+      "loss": 0.8916,
+      "step": 7687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.241409980659609,
+      "learning_rate": 3.987489035855574e-05,
+      "loss": 1.4492,
+      "step": 7688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8842668870562314,
+      "learning_rate": 3.987153815903397e-05,
+      "loss": 1.2754,
+      "step": 7689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1427805273800877,
+      "learning_rate": 3.986818554564334e-05,
+      "loss": 0.7866,
+      "step": 7690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7646999212852448,
+      "learning_rate": 3.9864832518477174e-05,
+      "loss": 1.0996,
+      "step": 7691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.332751724141572,
+      "learning_rate": 3.986147907762876e-05,
+      "loss": 1.7715,
+      "step": 7692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.945106152512718,
+      "learning_rate": 3.985812522319143e-05,
+      "loss": 1.0674,
+      "step": 7693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4282866238840866,
+      "learning_rate": 3.985477095525854e-05,
+      "loss": 1.3398,
+      "step": 7694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.350768806717249,
+      "learning_rate": 3.985141627392341e-05,
+      "loss": 1.4199,
+      "step": 7695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.354950301915114,
+      "learning_rate": 3.984806117927942e-05,
+      "loss": 0.8687,
+      "step": 7696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7543567478379325,
+      "learning_rate": 3.984470567141991e-05,
+      "loss": 1.0703,
+      "step": 7697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.028044588735785,
+      "learning_rate": 3.984134975043831e-05,
+      "loss": 1.0273,
+      "step": 7698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8774213131536914,
+      "learning_rate": 3.9837993416427975e-05,
+      "loss": 1.3916,
+      "step": 7699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.1213290798335,
+      "learning_rate": 3.983463666948233e-05,
+      "loss": 1.2422,
+      "step": 7700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.634864503641921,
+      "learning_rate": 3.983127950969477e-05,
+      "loss": 0.939,
+      "step": 7701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.363486395592734,
+      "learning_rate": 3.982792193715875e-05,
+      "loss": 0.7339,
+      "step": 7702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.391473479396558,
+      "learning_rate": 3.982456395196771e-05,
+      "loss": 1.042,
+      "step": 7703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3431341760943605,
+      "learning_rate": 3.982120555421508e-05,
+      "loss": 0.9922,
+      "step": 7704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.040917017085764,
+      "learning_rate": 3.981784674399433e-05,
+      "loss": 2.0391,
+      "step": 7705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9680919455705017,
+      "learning_rate": 3.981448752139894e-05,
+      "loss": 1.4258,
+      "step": 7706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.64617580289978,
+      "learning_rate": 3.981112788652239e-05,
+      "loss": 0.9414,
+      "step": 7707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2108024852492316,
+      "learning_rate": 3.980776783945819e-05,
+      "loss": 0.9766,
+      "step": 7708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.41183932256,
+      "learning_rate": 3.980440738029984e-05,
+      "loss": 0.7007,
+      "step": 7709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.618915449982032,
+      "learning_rate": 3.980104650914085e-05,
+      "loss": 0.9463,
+      "step": 7710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.011518538227939,
+      "learning_rate": 3.979768522607476e-05,
+      "loss": 1.0156,
+      "step": 7711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0980408385223157,
+      "learning_rate": 3.979432353119512e-05,
+      "loss": 0.8181,
+      "step": 7712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.0386837573272825,
+      "learning_rate": 3.979096142459548e-05,
+      "loss": 0.4619,
+      "step": 7713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9675537911788625,
+      "learning_rate": 3.978759890636941e-05,
+      "loss": 1.0791,
+      "step": 7714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.446004458868621,
+      "learning_rate": 3.978423597661047e-05,
+      "loss": 1.252,
+      "step": 7715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1408872791748212,
+      "learning_rate": 3.9780872635412256e-05,
+      "loss": 0.6443,
+      "step": 7716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.461286731947656,
+      "learning_rate": 3.977750888286839e-05,
+      "loss": 0.8135,
+      "step": 7717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.313091099430616,
+      "learning_rate": 3.977414471907246e-05,
+      "loss": 1.0371,
+      "step": 7718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.510390894627523,
+      "learning_rate": 3.97707801441181e-05,
+      "loss": 1.5254,
+      "step": 7719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.460263598981001,
+      "learning_rate": 3.976741515809894e-05,
+      "loss": 1.0928,
+      "step": 7720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.785737992538036,
+      "learning_rate": 3.976404976110863e-05,
+      "loss": 1.1953,
+      "step": 7721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.172952102660292,
+      "learning_rate": 3.976068395324083e-05,
+      "loss": 1.1484,
+      "step": 7722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.055821489878921,
+      "learning_rate": 3.9757317734589194e-05,
+      "loss": 1.3799,
+      "step": 7723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2020480694852838,
+      "learning_rate": 3.975395110524742e-05,
+      "loss": 0.833,
+      "step": 7724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1421247365982206,
+      "learning_rate": 3.9750584065309196e-05,
+      "loss": 1.0918,
+      "step": 7725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0799172697474795,
+      "learning_rate": 3.974721661486822e-05,
+      "loss": 1.0527,
+      "step": 7726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.90208875478077,
+      "learning_rate": 3.97438487540182e-05,
+      "loss": 1.0986,
+      "step": 7727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.752181228486746,
+      "learning_rate": 3.974048048285288e-05,
+      "loss": 0.6953,
+      "step": 7728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.532000647003511,
+      "learning_rate": 3.973711180146599e-05,
+      "loss": 1.0781,
+      "step": 7729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1003334050703666,
+      "learning_rate": 3.973374270995127e-05,
+      "loss": 1.123,
+      "step": 7730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0074380220415016,
+      "learning_rate": 3.97303732084025e-05,
+      "loss": 0.9541,
+      "step": 7731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.453095874356507,
+      "learning_rate": 3.9727003296913436e-05,
+      "loss": 1.2461,
+      "step": 7732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.995688798591292,
+      "learning_rate": 3.972363297557786e-05,
+      "loss": 0.8516,
+      "step": 7733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.958193870867646,
+      "learning_rate": 3.972026224448957e-05,
+      "loss": 1.6094,
+      "step": 7734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.177482227730297,
+      "learning_rate": 3.971689110374238e-05,
+      "loss": 1.5566,
+      "step": 7735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4464724778421005,
+      "learning_rate": 3.9713519553430115e-05,
+      "loss": 0.9043,
+      "step": 7736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.077262465658445,
+      "learning_rate": 3.971014759364657e-05,
+      "loss": 1.6445,
+      "step": 7737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.988481910165456,
+      "learning_rate": 3.9706775224485614e-05,
+      "loss": 1.8555,
+      "step": 7738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4405660663674316,
+      "learning_rate": 3.9703402446041094e-05,
+      "loss": 1.0664,
+      "step": 7739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8742348930821042,
+      "learning_rate": 3.970002925840687e-05,
+      "loss": 0.9263,
+      "step": 7740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.523804486256232,
+      "learning_rate": 3.969665566167681e-05,
+      "loss": 1.2336,
+      "step": 7741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3778197200859266,
+      "learning_rate": 3.9693281655944804e-05,
+      "loss": 0.8989,
+      "step": 7742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0813195477570203,
+      "learning_rate": 3.968990724130476e-05,
+      "loss": 0.957,
+      "step": 7743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6256431518304986,
+      "learning_rate": 3.968653241785057e-05,
+      "loss": 1.3037,
+      "step": 7744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9883491592528957,
+      "learning_rate": 3.968315718567617e-05,
+      "loss": 1.5022,
+      "step": 7745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3923232927733866,
+      "learning_rate": 3.967978154487547e-05,
+      "loss": 1.2656,
+      "step": 7746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2766195883102496,
+      "learning_rate": 3.967640549554244e-05,
+      "loss": 1.1318,
+      "step": 7747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0329371211033056,
+      "learning_rate": 3.967302903777101e-05,
+      "loss": 1.0015,
+      "step": 7748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5949820355429405,
+      "learning_rate": 3.9669652171655156e-05,
+      "loss": 1.1963,
+      "step": 7749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9757504208513748,
+      "learning_rate": 3.966627489728886e-05,
+      "loss": 1.4346,
+      "step": 7750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.025502210289289,
+      "learning_rate": 3.9662897214766106e-05,
+      "loss": 1.752,
+      "step": 7751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.783864795936916,
+      "learning_rate": 3.9659519124180885e-05,
+      "loss": 1.541,
+      "step": 7752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.81593570667261,
+      "learning_rate": 3.965614062562721e-05,
+      "loss": 0.7051,
+      "step": 7753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.548177106631417,
+      "learning_rate": 3.9652761719199115e-05,
+      "loss": 0.7102,
+      "step": 7754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.460936046781565,
+      "learning_rate": 3.9649382404990634e-05,
+      "loss": 1.0029,
+      "step": 7755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.110749199725791,
+      "learning_rate": 3.96460026830958e-05,
+      "loss": 0.757,
+      "step": 7756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7167024465751743,
+      "learning_rate": 3.9642622553608665e-05,
+      "loss": 0.3936,
+      "step": 7757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.531287675741989,
+      "learning_rate": 3.963924201662331e-05,
+      "loss": 1.1797,
+      "step": 7758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.181027345852342,
+      "learning_rate": 3.9635861072233815e-05,
+      "loss": 0.6611,
+      "step": 7759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5364056111103888,
+      "learning_rate": 3.9632479720534255e-05,
+      "loss": 0.8828,
+      "step": 7760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.356876901424156,
+      "learning_rate": 3.962909796161875e-05,
+      "loss": 1.7148,
+      "step": 7761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9592882862700085,
+      "learning_rate": 3.96257157955814e-05,
+      "loss": 1.1807,
+      "step": 7762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6382834080084914,
+      "learning_rate": 3.9622333222516326e-05,
+      "loss": 0.6555,
+      "step": 7763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.18349103426613,
+      "learning_rate": 3.961895024251768e-05,
+      "loss": 0.9854,
+      "step": 7764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9428244616560715,
+      "learning_rate": 3.96155668556796e-05,
+      "loss": 0.6057,
+      "step": 7765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7033830383792434,
+      "learning_rate": 3.961218306209625e-05,
+      "loss": 1.4609,
+      "step": 7766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.69088968423508,
+      "learning_rate": 3.9608798861861784e-05,
+      "loss": 0.9268,
+      "step": 7767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1849475683210926,
+      "learning_rate": 3.9605414255070396e-05,
+      "loss": 0.6724,
+      "step": 7768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.627007027483817,
+      "learning_rate": 3.9602029241816276e-05,
+      "loss": 0.4487,
+      "step": 7769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6266527876350105,
+      "learning_rate": 3.9598643822193626e-05,
+      "loss": 1.2402,
+      "step": 7770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6086939305493297,
+      "learning_rate": 3.959525799629665e-05,
+      "loss": 1.1553,
+      "step": 7771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.040964476020125,
+      "learning_rate": 3.9591871764219603e-05,
+      "loss": 1.1592,
+      "step": 7772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0648298159673213,
+      "learning_rate": 3.95884851260567e-05,
+      "loss": 1.2773,
+      "step": 7773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9995945281043732,
+      "learning_rate": 3.9585098081902193e-05,
+      "loss": 0.791,
+      "step": 7774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2376369668273934,
+      "learning_rate": 3.958171063185035e-05,
+      "loss": 0.8525,
+      "step": 7775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.416392431897671,
+      "learning_rate": 3.957832277599543e-05,
+      "loss": 0.3423,
+      "step": 7776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9832897509866996,
+      "learning_rate": 3.9574934514431724e-05,
+      "loss": 0.8271,
+      "step": 7777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4111114247915957,
+      "learning_rate": 3.957154584725353e-05,
+      "loss": 1.4902,
+      "step": 7778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3238073674139104,
+      "learning_rate": 3.9568156774555135e-05,
+      "loss": 1.2346,
+      "step": 7779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.894777851179296,
+      "learning_rate": 3.9564767296430877e-05,
+      "loss": 1.0371,
+      "step": 7780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.296397943034305,
+      "learning_rate": 3.956137741297507e-05,
+      "loss": 1.5684,
+      "step": 7781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4392636105641614,
+      "learning_rate": 3.955798712428206e-05,
+      "loss": 0.9707,
+      "step": 7782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.0780627951134685,
+      "learning_rate": 3.955459643044619e-05,
+      "loss": 0.5464,
+      "step": 7783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.479571995872442,
+      "learning_rate": 3.9551205331561834e-05,
+      "loss": 1.1465,
+      "step": 7784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.272424441808411,
+      "learning_rate": 3.9547813827723344e-05,
+      "loss": 0.9526,
+      "step": 7785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2354175895956523,
+      "learning_rate": 3.9544421919025135e-05,
+      "loss": 0.999,
+      "step": 7786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.477157438999681,
+      "learning_rate": 3.954102960556157e-05,
+      "loss": 1.1953,
+      "step": 7787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.111715191306426,
+      "learning_rate": 3.953763688742708e-05,
+      "loss": 0.9407,
+      "step": 7788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8338341176892463,
+      "learning_rate": 3.953424376471607e-05,
+      "loss": 1.3984,
+      "step": 7789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7349071889013437,
+      "learning_rate": 3.9530850237522974e-05,
+      "loss": 0.7638,
+      "step": 7790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.926184543070304,
+      "learning_rate": 3.952745630594223e-05,
+      "loss": 1.3711,
+      "step": 7791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3232360300990282,
+      "learning_rate": 3.9524061970068294e-05,
+      "loss": 1.0322,
+      "step": 7792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.005380111734408,
+      "learning_rate": 3.952066722999563e-05,
+      "loss": 1.1641,
+      "step": 7793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7277296625687364,
+      "learning_rate": 3.95172720858187e-05,
+      "loss": 1.2568,
+      "step": 7794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0198672105558844,
+      "learning_rate": 3.9513876537632e-05,
+      "loss": 1.0156,
+      "step": 7795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.774094560991149,
+      "learning_rate": 3.951048058553003e-05,
+      "loss": 1.6113,
+      "step": 7796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7810172508662148,
+      "learning_rate": 3.950708422960729e-05,
+      "loss": 1.0518,
+      "step": 7797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0759036015526955,
+      "learning_rate": 3.95036874699583e-05,
+      "loss": 0.5339,
+      "step": 7798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.365091140226404,
+      "learning_rate": 3.950029030667759e-05,
+      "loss": 1.2983,
+      "step": 7799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.748426708376232,
+      "learning_rate": 3.94968927398597e-05,
+      "loss": 1.7476,
+      "step": 7800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4159069730709746,
+      "learning_rate": 3.9493494769599195e-05,
+      "loss": 0.8955,
+      "step": 7801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.422309295108078,
+      "learning_rate": 3.9490096395990626e-05,
+      "loss": 1.1406,
+      "step": 7802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.9318252982130995,
+      "learning_rate": 3.948669761912858e-05,
+      "loss": 0.6202,
+      "step": 7803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9062379930462021,
+      "learning_rate": 3.948329843910763e-05,
+      "loss": 1.4746,
+      "step": 7804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.528089802501877,
+      "learning_rate": 3.947989885602238e-05,
+      "loss": 1.3418,
+      "step": 7805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4705614738345365,
+      "learning_rate": 3.947649886996745e-05,
+      "loss": 0.9727,
+      "step": 7806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.268173955006722,
+      "learning_rate": 3.947309848103744e-05,
+      "loss": 1.1455,
+      "step": 7807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.803430904656644,
+      "learning_rate": 3.946969768932699e-05,
+      "loss": 0.7964,
+      "step": 7808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8846231372435636,
+      "learning_rate": 3.946629649493076e-05,
+      "loss": 0.9512,
+      "step": 7809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.349726224989879,
+      "learning_rate": 3.946289489794337e-05,
+      "loss": 1.3066,
+      "step": 7810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.507195977149174,
+      "learning_rate": 3.9459492898459504e-05,
+      "loss": 1.251,
+      "step": 7811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6430534617587833,
+      "learning_rate": 3.945609049657385e-05,
+      "loss": 0.8828,
+      "step": 7812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9233575823251905,
+      "learning_rate": 3.945268769238106e-05,
+      "loss": 0.7271,
+      "step": 7813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.8896920286216,
+      "learning_rate": 3.944928448597588e-05,
+      "loss": 0.75,
+      "step": 7814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9611709881263741,
+      "learning_rate": 3.944588087745298e-05,
+      "loss": 0.8213,
+      "step": 7815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.465736523580364,
+      "learning_rate": 3.94424768669071e-05,
+      "loss": 0.7749,
+      "step": 7816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.820861847440851,
+      "learning_rate": 3.9439072454432976e-05,
+      "loss": 1.0977,
+      "step": 7817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.080851780824032,
+      "learning_rate": 3.943566764012533e-05,
+      "loss": 0.876,
+      "step": 7818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.858622791653621,
+      "learning_rate": 3.943226242407894e-05,
+      "loss": 1.7549,
+      "step": 7819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.783019891409214,
+      "learning_rate": 3.942885680638856e-05,
+      "loss": 1.4795,
+      "step": 7820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.484188528680263,
+      "learning_rate": 3.9425450787148965e-05,
+      "loss": 1.3945,
+      "step": 7821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.301420631361167,
+      "learning_rate": 3.9422044366454944e-05,
+      "loss": 1.2148,
+      "step": 7822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.531604621427855,
+      "learning_rate": 3.941863754440131e-05,
+      "loss": 0.8936,
+      "step": 7823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.530502783892039,
+      "learning_rate": 3.941523032108286e-05,
+      "loss": 1.236,
+      "step": 7824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.119658673762835,
+      "learning_rate": 3.9411822696594415e-05,
+      "loss": 1.6543,
+      "step": 7825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.690265360512361,
+      "learning_rate": 3.9408414671030816e-05,
+      "loss": 0.7178,
+      "step": 7826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.027092774001973,
+      "learning_rate": 3.9405006244486895e-05,
+      "loss": 1.1221,
+      "step": 7827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7104387071804594,
+      "learning_rate": 3.9401597417057526e-05,
+      "loss": 0.9014,
+      "step": 7828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.397371911747357,
+      "learning_rate": 3.939818818883756e-05,
+      "loss": 1.1816,
+      "step": 7829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.769666473651323,
+      "learning_rate": 3.939477855992188e-05,
+      "loss": 1.2812,
+      "step": 7830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9091151858309612,
+      "learning_rate": 3.939136853040536e-05,
+      "loss": 1.1895,
+      "step": 7831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.732061224607469,
+      "learning_rate": 3.938795810038291e-05,
+      "loss": 1.0059,
+      "step": 7832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.498701998396033,
+      "learning_rate": 3.9384547269949465e-05,
+      "loss": 1.168,
+      "step": 7833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.216548615278727,
+      "learning_rate": 3.938113603919992e-05,
+      "loss": 1.8828,
+      "step": 7834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9561817277799833,
+      "learning_rate": 3.93777244082292e-05,
+      "loss": 0.9189,
+      "step": 7835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5170388365497476,
+      "learning_rate": 3.937431237713227e-05,
+      "loss": 0.593,
+      "step": 7836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.055320146483939,
+      "learning_rate": 3.937089994600408e-05,
+      "loss": 1.1025,
+      "step": 7837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8797549856126032,
+      "learning_rate": 3.9367487114939596e-05,
+      "loss": 1.0078,
+      "step": 7838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5351004807461277,
+      "learning_rate": 3.936407388403379e-05,
+      "loss": 1.3281,
+      "step": 7839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.263876197653595,
+      "learning_rate": 3.936066025338167e-05,
+      "loss": 1.5869,
+      "step": 7840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.938275498635856,
+      "learning_rate": 3.9357246223078204e-05,
+      "loss": 1.124,
+      "step": 7841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3140426498458884,
+      "learning_rate": 3.9353831793218424e-05,
+      "loss": 1.9316,
+      "step": 7842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8929079632478887,
+      "learning_rate": 3.935041696389735e-05,
+      "loss": 0.9832,
+      "step": 7843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.625909458130296,
+      "learning_rate": 3.934700173521001e-05,
+      "loss": 1.2354,
+      "step": 7844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.175129991408432,
+      "learning_rate": 3.934358610725146e-05,
+      "loss": 1.1465,
+      "step": 7845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.107167196677128,
+      "learning_rate": 3.9340170080116745e-05,
+      "loss": 1.3301,
+      "step": 7846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.194931723365267,
+      "learning_rate": 3.9336753653900935e-05,
+      "loss": 0.8311,
+      "step": 7847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9773439998421405,
+      "learning_rate": 3.9333336828699106e-05,
+      "loss": 1.1504,
+      "step": 7848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.222583978336199,
+      "learning_rate": 3.932991960460636e-05,
+      "loss": 1.3828,
+      "step": 7849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.877931178687686,
+      "learning_rate": 3.9326501981717765e-05,
+      "loss": 1.0811,
+      "step": 7850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9984816841658914,
+      "learning_rate": 3.932308396012846e-05,
+      "loss": 0.6675,
+      "step": 7851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9494908491510996,
+      "learning_rate": 3.931966553993356e-05,
+      "loss": 0.3884,
+      "step": 7852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5140220321138997,
+      "learning_rate": 3.93162467212282e-05,
+      "loss": 0.8779,
+      "step": 7853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.539727320564845,
+      "learning_rate": 3.9312827504107526e-05,
+      "loss": 1.0605,
+      "step": 7854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6450550005842453,
+      "learning_rate": 3.930940788866668e-05,
+      "loss": 1.2568,
+      "step": 7855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.589209272212244,
+      "learning_rate": 3.930598787500084e-05,
+      "loss": 0.863,
+      "step": 7856
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.1257936118832275,
+      "learning_rate": 3.930256746320519e-05,
+      "loss": 1.0215,
+      "step": 7857
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9241562246163584,
+      "learning_rate": 3.9299146653374893e-05,
+      "loss": 1.1797,
+      "step": 7858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.540797277711664,
+      "learning_rate": 3.929572544560518e-05,
+      "loss": 0.8628,
+      "step": 7859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4638979588967294,
+      "learning_rate": 3.929230383999124e-05,
+      "loss": 1.1123,
+      "step": 7860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3216294493044116,
+      "learning_rate": 3.928888183662831e-05,
+      "loss": 1.2168,
+      "step": 7861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.595841172535036,
+      "learning_rate": 3.928545943561161e-05,
+      "loss": 1.6504,
+      "step": 7862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3435123577439136,
+      "learning_rate": 3.928203663703639e-05,
+      "loss": 0.8184,
+      "step": 7863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4577702136584882,
+      "learning_rate": 3.92786134409979e-05,
+      "loss": 1.2812,
+      "step": 7864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.153109810738703,
+      "learning_rate": 3.927518984759142e-05,
+      "loss": 0.9824,
+      "step": 7865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.807526433577448,
+      "learning_rate": 3.927176585691222e-05,
+      "loss": 1.5312,
+      "step": 7866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6325385519974223,
+      "learning_rate": 3.926834146905558e-05,
+      "loss": 1.0469,
+      "step": 7867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.07929561327178,
+      "learning_rate": 3.9264916684116807e-05,
+      "loss": 0.9736,
+      "step": 7868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.574321950893356,
+      "learning_rate": 3.92614915021912e-05,
+      "loss": 1.0371,
+      "step": 7869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7378861037445725,
+      "learning_rate": 3.92580659233741e-05,
+      "loss": 0.8706,
+      "step": 7870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.666562285116393,
+      "learning_rate": 3.925463994776083e-05,
+      "loss": 1.5977,
+      "step": 7871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6367414332418933,
+      "learning_rate": 3.925121357544674e-05,
+      "loss": 0.9541,
+      "step": 7872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9786739519907286,
+      "learning_rate": 3.924778680652716e-05,
+      "loss": 1.123,
+      "step": 7873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2563360514042916,
+      "learning_rate": 3.924435964109749e-05,
+      "loss": 0.7803,
+      "step": 7874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.295131722453656,
+      "learning_rate": 3.924093207925308e-05,
+      "loss": 1.2207,
+      "step": 7875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.723536608908956,
+      "learning_rate": 3.923750412108933e-05,
+      "loss": 0.9561,
+      "step": 7876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0081501560677113,
+      "learning_rate": 3.9234075766701646e-05,
+      "loss": 1.0801,
+      "step": 7877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5714626337879354,
+      "learning_rate": 3.9230647016185416e-05,
+      "loss": 0.8291,
+      "step": 7878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3345119458921366,
+      "learning_rate": 3.922721786963608e-05,
+      "loss": 0.54,
+      "step": 7879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.17756587860624,
+      "learning_rate": 3.922378832714906e-05,
+      "loss": 1.1406,
+      "step": 7880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.341587990088657,
+      "learning_rate": 3.92203583888198e-05,
+      "loss": 1.3848,
+      "step": 7881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.044941352394257,
+      "learning_rate": 3.921692805474375e-05,
+      "loss": 0.9795,
+      "step": 7882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.768786972922489,
+      "learning_rate": 3.921349732501639e-05,
+      "loss": 1.7793,
+      "step": 7883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.205616815810362,
+      "learning_rate": 3.921006619973318e-05,
+      "loss": 1.1289,
+      "step": 7884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.665409190401674,
+      "learning_rate": 3.920663467898962e-05,
+      "loss": 0.717,
+      "step": 7885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.800292435486359,
+      "learning_rate": 3.920320276288119e-05,
+      "loss": 0.833,
+      "step": 7886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8721460238545495,
+      "learning_rate": 3.919977045150341e-05,
+      "loss": 1.083,
+      "step": 7887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0851807257888164,
+      "learning_rate": 3.919633774495181e-05,
+      "loss": 0.6172,
+      "step": 7888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6107132911981141,
+      "learning_rate": 3.9192904643321896e-05,
+      "loss": 0.7617,
+      "step": 7889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.706399951330867,
+      "learning_rate": 3.918947114670923e-05,
+      "loss": 1.168,
+      "step": 7890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6230965249814129,
+      "learning_rate": 3.918603725520936e-05,
+      "loss": 0.2531,
+      "step": 7891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.458083382433318,
+      "learning_rate": 3.918260296891785e-05,
+      "loss": 1.8516,
+      "step": 7892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5885359970634485,
+      "learning_rate": 3.917916828793027e-05,
+      "loss": 0.8535,
+      "step": 7893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.156788726539977,
+      "learning_rate": 3.917573321234222e-05,
+      "loss": 1.3379,
+      "step": 7894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0937528467870687,
+      "learning_rate": 3.917229774224926e-05,
+      "loss": 1.3965,
+      "step": 7895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.138572955801516,
+      "learning_rate": 3.916886187774704e-05,
+      "loss": 1.4219,
+      "step": 7896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.733786971333397,
+      "learning_rate": 3.916542561893116e-05,
+      "loss": 0.9941,
+      "step": 7897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5867373835553877,
+      "learning_rate": 3.916198896589725e-05,
+      "loss": 1.1113,
+      "step": 7898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.473651890717198,
+      "learning_rate": 3.915855191874096e-05,
+      "loss": 1.1934,
+      "step": 7899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6220428522188706,
+      "learning_rate": 3.9155114477557933e-05,
+      "loss": 1.6699,
+      "step": 7900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.052689076188517,
+      "learning_rate": 3.915167664244382e-05,
+      "loss": 1.4551,
+      "step": 7901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2581340204476303,
+      "learning_rate": 3.914823841349432e-05,
+      "loss": 1.1641,
+      "step": 7902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3864298238564308,
+      "learning_rate": 3.91447997908051e-05,
+      "loss": 0.8418,
+      "step": 7903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6003951469337374,
+      "learning_rate": 3.9141360774471865e-05,
+      "loss": 0.4175,
+      "step": 7904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.986771632193067,
+      "learning_rate": 3.913792136459031e-05,
+      "loss": 1.1172,
+      "step": 7905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3156274391043103,
+      "learning_rate": 3.913448156125616e-05,
+      "loss": 1.4316,
+      "step": 7906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5184035978817474,
+      "learning_rate": 3.913104136456515e-05,
+      "loss": 1.3496,
+      "step": 7907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5299688794178388,
+      "learning_rate": 3.9127600774613e-05,
+      "loss": 1.582,
+      "step": 7908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.819300219124223,
+      "learning_rate": 3.912415979149547e-05,
+      "loss": 1.0186,
+      "step": 7909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.293063383994481,
+      "learning_rate": 3.9120718415308334e-05,
+      "loss": 1.4531,
+      "step": 7910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.77889220347057,
+      "learning_rate": 3.911727664614735e-05,
+      "loss": 0.9541,
+      "step": 7911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2902692089843244,
+      "learning_rate": 3.91138344841083e-05,
+      "loss": 0.7168,
+      "step": 7912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.52103148290733,
+      "learning_rate": 3.911039192928698e-05,
+      "loss": 1.3682,
+      "step": 7913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7976803605724623,
+      "learning_rate": 3.9106948981779214e-05,
+      "loss": 1.1025,
+      "step": 7914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7800369188889817,
+      "learning_rate": 3.910350564168078e-05,
+      "loss": 0.8818,
+      "step": 7915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6295326699582815,
+      "learning_rate": 3.910006190908754e-05,
+      "loss": 0.7361,
+      "step": 7916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.673792212826929,
+      "learning_rate": 3.909661778409531e-05,
+      "loss": 1.1592,
+      "step": 7917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.331956400108547,
+      "learning_rate": 3.909317326679994e-05,
+      "loss": 1.3242,
+      "step": 7918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.372994317390175,
+      "learning_rate": 3.9089728357297304e-05,
+      "loss": 0.606,
+      "step": 7919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5186475179868415,
+      "learning_rate": 3.908628305568327e-05,
+      "loss": 1.0557,
+      "step": 7920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.240487760333745,
+      "learning_rate": 3.90828373620537e-05,
+      "loss": 1.1543,
+      "step": 7921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.3744240517938024,
+      "learning_rate": 3.9079391276504506e-05,
+      "loss": 0.7422,
+      "step": 7922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.704815275761151,
+      "learning_rate": 3.9075944799131584e-05,
+      "loss": 1.4497,
+      "step": 7923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.806641021815302,
+      "learning_rate": 3.907249793003084e-05,
+      "loss": 0.5688,
+      "step": 7924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.756956664687415,
+      "learning_rate": 3.9069050669298226e-05,
+      "loss": 0.5933,
+      "step": 7925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.73023972401819,
+      "learning_rate": 3.906560301702964e-05,
+      "loss": 0.8945,
+      "step": 7926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8770340068399736,
+      "learning_rate": 3.906215497332106e-05,
+      "loss": 0.8384,
+      "step": 7927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.397941776375749,
+      "learning_rate": 3.905870653826843e-05,
+      "loss": 1.2021,
+      "step": 7928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.633559939699383,
+      "learning_rate": 3.905525771196772e-05,
+      "loss": 1.6387,
+      "step": 7929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.445518606841643,
+      "learning_rate": 3.905180849451491e-05,
+      "loss": 1.7324,
+      "step": 7930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.631026028361091,
+      "learning_rate": 3.9048358886005986e-05,
+      "loss": 0.8555,
+      "step": 7931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7831388767255425,
+      "learning_rate": 3.9044908886536956e-05,
+      "loss": 0.7793,
+      "step": 7932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.523278742558957,
+      "learning_rate": 3.9041458496203834e-05,
+      "loss": 0.9639,
+      "step": 7933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7258519337073044,
+      "learning_rate": 3.903800771510263e-05,
+      "loss": 1.1348,
+      "step": 7934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.2015797577392755,
+      "learning_rate": 3.9034556543329386e-05,
+      "loss": 1.251,
+      "step": 7935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.881449223749834,
+      "learning_rate": 3.903110498098015e-05,
+      "loss": 1.6758,
+      "step": 7936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.711964560124486,
+      "learning_rate": 3.902765302815098e-05,
+      "loss": 0.7197,
+      "step": 7937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.185705266127155,
+      "learning_rate": 3.902420068493793e-05,
+      "loss": 0.6851,
+      "step": 7938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.700325049870774,
+      "learning_rate": 3.902074795143709e-05,
+      "loss": 1.582,
+      "step": 7939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.403949049443332,
+      "learning_rate": 3.901729482774453e-05,
+      "loss": 0.8857,
+      "step": 7940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.419443778181315,
+      "learning_rate": 3.901384131395638e-05,
+      "loss": 1.375,
+      "step": 7941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7901241700563597,
+      "learning_rate": 3.901038741016871e-05,
+      "loss": 1.0626,
+      "step": 7942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.309656947527389,
+      "learning_rate": 3.900693311647768e-05,
+      "loss": 0.9238,
+      "step": 7943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6968364443183357,
+      "learning_rate": 3.90034784329794e-05,
+      "loss": 1.1099,
+      "step": 7944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1398923413656816,
+      "learning_rate": 3.900002335977001e-05,
+      "loss": 0.9873,
+      "step": 7945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.504097346543376,
+      "learning_rate": 3.8996567896945666e-05,
+      "loss": 0.228,
+      "step": 7946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0652223595705657,
+      "learning_rate": 3.899311204460254e-05,
+      "loss": 0.9438,
+      "step": 7947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.448673883519665,
+      "learning_rate": 3.898965580283681e-05,
+      "loss": 1.1172,
+      "step": 7948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.576681987188068,
+      "learning_rate": 3.8986199171744644e-05,
+      "loss": 1.0664,
+      "step": 7949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9527309901670638,
+      "learning_rate": 3.898274215142226e-05,
+      "loss": 1.1211,
+      "step": 7950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.773636484447106,
+      "learning_rate": 3.8979284741965846e-05,
+      "loss": 1.5547,
+      "step": 7951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1408318189964812,
+      "learning_rate": 3.8975826943471624e-05,
+      "loss": 1.0137,
+      "step": 7952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0503934450120624,
+      "learning_rate": 3.897236875603584e-05,
+      "loss": 1.3438,
+      "step": 7953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4576280411886313,
+      "learning_rate": 3.8968910179754716e-05,
+      "loss": 1.1855,
+      "step": 7954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4053126899963995,
+      "learning_rate": 3.896545121472451e-05,
+      "loss": 1.1895,
+      "step": 7955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.723604403641534,
+      "learning_rate": 3.896199186104148e-05,
+      "loss": 1.2832,
+      "step": 7956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.852393112908536,
+      "learning_rate": 3.895853211880189e-05,
+      "loss": 0.8633,
+      "step": 7957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.93732685037415,
+      "learning_rate": 3.8955071988102046e-05,
+      "loss": 1.2734,
+      "step": 7958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3606062650947583,
+      "learning_rate": 3.895161146903823e-05,
+      "loss": 1.1475,
+      "step": 7959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.575597280067417,
+      "learning_rate": 3.8948150561706745e-05,
+      "loss": 1.0791,
+      "step": 7960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.276929762100827,
+      "learning_rate": 3.8944689266203905e-05,
+      "loss": 0.7817,
+      "step": 7961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.359977843964746,
+      "learning_rate": 3.894122758262604e-05,
+      "loss": 1.2458,
+      "step": 7962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.73950823490326,
+      "learning_rate": 3.893776551106949e-05,
+      "loss": 1.209,
+      "step": 7963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.343124611387223,
+      "learning_rate": 3.893430305163061e-05,
+      "loss": 1.2461,
+      "step": 7964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.834296075809174,
+      "learning_rate": 3.8930840204405726e-05,
+      "loss": 1.0801,
+      "step": 7965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.322147212533726,
+      "learning_rate": 3.892737696949125e-05,
+      "loss": 1.2715,
+      "step": 7966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.535105150511539,
+      "learning_rate": 3.8923913346983534e-05,
+      "loss": 0.6367,
+      "step": 7967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.540929019749889,
+      "learning_rate": 3.892044933697897e-05,
+      "loss": 1.5,
+      "step": 7968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.326053685411853,
+      "learning_rate": 3.891698493957398e-05,
+      "loss": 0.8418,
+      "step": 7969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6878624430839184,
+      "learning_rate": 3.8913520154864966e-05,
+      "loss": 0.8545,
+      "step": 7970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7827254250133313,
+      "learning_rate": 3.891005498294834e-05,
+      "loss": 0.6321,
+      "step": 7971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.815436546801608,
+      "learning_rate": 3.890658942392055e-05,
+      "loss": 0.748,
+      "step": 7972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.538383416549849,
+      "learning_rate": 3.8903123477878045e-05,
+      "loss": 1.8613,
+      "step": 7973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8409050764575685,
+      "learning_rate": 3.8899657144917266e-05,
+      "loss": 1.0957,
+      "step": 7974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.030112361841655,
+      "learning_rate": 3.889619042513469e-05,
+      "loss": 1.0527,
+      "step": 7975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8346865919076167,
+      "learning_rate": 3.889272331862679e-05,
+      "loss": 1.1836,
+      "step": 7976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.618584747700977,
+      "learning_rate": 3.888925582549006e-05,
+      "loss": 1.8672,
+      "step": 7977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.382353991810907,
+      "learning_rate": 3.888578794582099e-05,
+      "loss": 0.6421,
+      "step": 7978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4870995787308487,
+      "learning_rate": 3.8882319679716094e-05,
+      "loss": 0.7461,
+      "step": 7979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.42635477507038,
+      "learning_rate": 3.88788510272719e-05,
+      "loss": 0.7061,
+      "step": 7980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.077310631816132,
+      "learning_rate": 3.8875381988584925e-05,
+      "loss": 1.0039,
+      "step": 7981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.202269367695379,
+      "learning_rate": 3.887191256375172e-05,
+      "loss": 1.0356,
+      "step": 7982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.50502496540972,
+      "learning_rate": 3.8868442752868834e-05,
+      "loss": 0.6667,
+      "step": 7983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.8467512123538095,
+      "learning_rate": 3.886497255603283e-05,
+      "loss": 1.2476,
+      "step": 7984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.66881040559134,
+      "learning_rate": 3.8861501973340295e-05,
+      "loss": 1.7344,
+      "step": 7985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2850694427956535,
+      "learning_rate": 3.88580310048878e-05,
+      "loss": 1.2852,
+      "step": 7986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0042423163077885,
+      "learning_rate": 3.885455965077193e-05,
+      "loss": 0.5876,
+      "step": 7987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.573869037952551,
+      "learning_rate": 3.8851087911089315e-05,
+      "loss": 1.0547,
+      "step": 7988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.801169521655255,
+      "learning_rate": 3.884761578593657e-05,
+      "loss": 1.0742,
+      "step": 7989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.67259103652662,
+      "learning_rate": 3.88441432754103e-05,
+      "loss": 1.4551,
+      "step": 7990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.600527778221801,
+      "learning_rate": 3.884067037960717e-05,
+      "loss": 0.9697,
+      "step": 7991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.981853758257156,
+      "learning_rate": 3.8837197098623814e-05,
+      "loss": 1.2539,
+      "step": 7992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.733309529868696,
+      "learning_rate": 3.883372343255689e-05,
+      "loss": 0.8975,
+      "step": 7993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.0036535266245155,
+      "learning_rate": 3.8830249381503084e-05,
+      "loss": 0.8154,
+      "step": 7994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.324182949271281,
+      "learning_rate": 3.8826774945559055e-05,
+      "loss": 1.1688,
+      "step": 7995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0834311048797636,
+      "learning_rate": 3.8823300124821524e-05,
+      "loss": 0.7979,
+      "step": 7996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.886367065557523,
+      "learning_rate": 3.881982491938717e-05,
+      "loss": 1.7812,
+      "step": 7997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8973803304296668,
+      "learning_rate": 3.8816349329352706e-05,
+      "loss": 0.8127,
+      "step": 7998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.920150073545794,
+      "learning_rate": 3.881287335481488e-05,
+      "loss": 0.5272,
+      "step": 7999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5609138045457023,
+      "learning_rate": 3.880939699587041e-05,
+      "loss": 1.3184,
+      "step": 8000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1542916734482196,
+      "learning_rate": 3.880592025261603e-05,
+      "loss": 0.9199,
+      "step": 8001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.257977012670269,
+      "learning_rate": 3.8802443125148526e-05,
+      "loss": 1.0996,
+      "step": 8002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.029418350938569,
+      "learning_rate": 3.879896561356464e-05,
+      "loss": 1.5234,
+      "step": 8003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.555104571920694,
+      "learning_rate": 3.879548771796117e-05,
+      "loss": 1.8438,
+      "step": 8004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2716678404251094,
+      "learning_rate": 3.8792009438434874e-05,
+      "loss": 1.1084,
+      "step": 8005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.212498952574401,
+      "learning_rate": 3.878853077508258e-05,
+      "loss": 0.7183,
+      "step": 8006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3568450297835204,
+      "learning_rate": 3.878505172800109e-05,
+      "loss": 1.2568,
+      "step": 8007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4739442095437023,
+      "learning_rate": 3.878157229728722e-05,
+      "loss": 0.9189,
+      "step": 8008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1711424935432135,
+      "learning_rate": 3.8778092483037796e-05,
+      "loss": 1.1729,
+      "step": 8009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9399310887308365,
+      "learning_rate": 3.877461228534968e-05,
+      "loss": 0.8735,
+      "step": 8010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7802705581650125,
+      "learning_rate": 3.877113170431971e-05,
+      "loss": 1.0137,
+      "step": 8011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.183135533817558,
+      "learning_rate": 3.876765074004474e-05,
+      "loss": 1.1016,
+      "step": 8012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.775918801590508,
+      "learning_rate": 3.8764169392621664e-05,
+      "loss": 1.1992,
+      "step": 8013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.016344065451227,
+      "learning_rate": 3.876068766214735e-05,
+      "loss": 1.5254,
+      "step": 8014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.776090451032364,
+      "learning_rate": 3.875720554871871e-05,
+      "loss": 1.6348,
+      "step": 8015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8602844896383455,
+      "learning_rate": 3.875372305243263e-05,
+      "loss": 0.8018,
+      "step": 8016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.154557331303702,
+      "learning_rate": 3.875024017338604e-05,
+      "loss": 1.2441,
+      "step": 8017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6119074610372346,
+      "learning_rate": 3.874675691167587e-05,
+      "loss": 1.6582,
+      "step": 8018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9010435660245877,
+      "learning_rate": 3.874327326739905e-05,
+      "loss": 0.7524,
+      "step": 8019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.891181997850148,
+      "learning_rate": 3.8739789240652524e-05,
+      "loss": 0.8174,
+      "step": 8020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.540562958094715,
+      "learning_rate": 3.873630483153326e-05,
+      "loss": 1.3594,
+      "step": 8021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.261970897316827,
+      "learning_rate": 3.8732820040138226e-05,
+      "loss": 1.2979,
+      "step": 8022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3092482102294145,
+      "learning_rate": 3.8729334866564404e-05,
+      "loss": 1.1875,
+      "step": 8023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5038078574970832,
+      "learning_rate": 3.872584931090878e-05,
+      "loss": 0.7656,
+      "step": 8024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.347663371947646,
+      "learning_rate": 3.872236337326836e-05,
+      "loss": 0.8276,
+      "step": 8025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1068247885253006,
+      "learning_rate": 3.871887705374014e-05,
+      "loss": 1.2266,
+      "step": 8026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5475297332016815,
+      "learning_rate": 3.871539035242117e-05,
+      "loss": 0.8823,
+      "step": 8027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2910704675274642,
+      "learning_rate": 3.871190326940846e-05,
+      "loss": 1.2285,
+      "step": 8028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4354797207186385,
+      "learning_rate": 3.870841580479907e-05,
+      "loss": 0.6755,
+      "step": 8029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.817698259772689,
+      "learning_rate": 3.870492795869005e-05,
+      "loss": 1.3945,
+      "step": 8030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.542337131141129,
+      "learning_rate": 3.870143973117846e-05,
+      "loss": 0.9326,
+      "step": 8031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7973494856931174,
+      "learning_rate": 3.869795112236138e-05,
+      "loss": 0.8765,
+      "step": 8032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8552915672093553,
+      "learning_rate": 3.86944621323359e-05,
+      "loss": 0.6572,
+      "step": 8033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.817406667251651,
+      "learning_rate": 3.869097276119911e-05,
+      "loss": 1.3418,
+      "step": 8034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.6846037816024255,
+      "learning_rate": 3.8687483009048127e-05,
+      "loss": 0.9302,
+      "step": 8035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9562685750172086,
+      "learning_rate": 3.868399287598006e-05,
+      "loss": 0.8291,
+      "step": 8036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2718832989190862,
+      "learning_rate": 3.868050236209204e-05,
+      "loss": 0.325,
+      "step": 8037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7073156254231607,
+      "learning_rate": 3.8677011467481217e-05,
+      "loss": 0.8984,
+      "step": 8038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7425032384905568,
+      "learning_rate": 3.867352019224473e-05,
+      "loss": 1.4037,
+      "step": 8039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3228936529231223,
+      "learning_rate": 3.867002853647974e-05,
+      "loss": 0.7363,
+      "step": 8040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6921367124614535,
+      "learning_rate": 3.866653650028342e-05,
+      "loss": 1.0625,
+      "step": 8041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4055749516026976,
+      "learning_rate": 3.866304408375295e-05,
+      "loss": 1.6973,
+      "step": 8042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2058268247346096,
+      "learning_rate": 3.8659551286985534e-05,
+      "loss": 0.9277,
+      "step": 8043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.728784351491505,
+      "learning_rate": 3.8656058110078365e-05,
+      "loss": 1.0103,
+      "step": 8044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.358020310976063,
+      "learning_rate": 3.8652564553128654e-05,
+      "loss": 1.3672,
+      "step": 8045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.171012432693594,
+      "learning_rate": 3.864907061623363e-05,
+      "loss": 1.1211,
+      "step": 8046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.287310761237345,
+      "learning_rate": 3.8645576299490524e-05,
+      "loss": 1.4238,
+      "step": 8047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2615804494444474,
+      "learning_rate": 3.864208160299659e-05,
+      "loss": 1.0811,
+      "step": 8048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.050819855859686,
+      "learning_rate": 3.8638586526849074e-05,
+      "loss": 1.0879,
+      "step": 8049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.84718180808254,
+      "learning_rate": 3.863509107114526e-05,
+      "loss": 1.5479,
+      "step": 8050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.330979463570989,
+      "learning_rate": 3.86315952359824e-05,
+      "loss": 1.2109,
+      "step": 8051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0277876469039553,
+      "learning_rate": 3.86280990214578e-05,
+      "loss": 1.0283,
+      "step": 8052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.315887238638229,
+      "learning_rate": 3.8624602427668746e-05,
+      "loss": 1.4277,
+      "step": 8053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.263598821930755,
+      "learning_rate": 3.862110545471256e-05,
+      "loss": 1.3643,
+      "step": 8054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.285244157527552,
+      "learning_rate": 3.8617608102686546e-05,
+      "loss": 1.1611,
+      "step": 8055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.311773546915973,
+      "learning_rate": 3.861411037168805e-05,
+      "loss": 1.4473,
+      "step": 8056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8598053821957263,
+      "learning_rate": 3.86106122618144e-05,
+      "loss": 0.6846,
+      "step": 8057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4211139268055817,
+      "learning_rate": 3.8607113773162954e-05,
+      "loss": 1.8516,
+      "step": 8058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6706227866045467,
+      "learning_rate": 3.860361490583106e-05,
+      "loss": 0.958,
+      "step": 8059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.681603697456483,
+      "learning_rate": 3.8600115659916116e-05,
+      "loss": 1.4004,
+      "step": 8060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7926720094767035,
+      "learning_rate": 3.859661603551549e-05,
+      "loss": 1.1016,
+      "step": 8061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.128013045493272,
+      "learning_rate": 3.8593116032726573e-05,
+      "loss": 1.127,
+      "step": 8062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.539774445709088,
+      "learning_rate": 3.858961565164676e-05,
+      "loss": 0.6294,
+      "step": 8063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2973841926685106,
+      "learning_rate": 3.858611489237349e-05,
+      "loss": 1.124,
+      "step": 8064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.289579157410395,
+      "learning_rate": 3.858261375500417e-05,
+      "loss": 1.25,
+      "step": 8065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.951412749856254,
+      "learning_rate": 3.857911223963623e-05,
+      "loss": 1.2549,
+      "step": 8066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7060658768884056,
+      "learning_rate": 3.857561034636713e-05,
+      "loss": 1.2119,
+      "step": 8067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0283203125,
+      "learning_rate": 3.857210807529432e-05,
+      "loss": 0.5825,
+      "step": 8068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.186662732586156,
+      "learning_rate": 3.856860542651527e-05,
+      "loss": 0.791,
+      "step": 8069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.226734380932166,
+      "learning_rate": 3.856510240012745e-05,
+      "loss": 1.6465,
+      "step": 8070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.743550019156036,
+      "learning_rate": 3.856159899622836e-05,
+      "loss": 0.8213,
+      "step": 8071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6559903186196516,
+      "learning_rate": 3.855809521491549e-05,
+      "loss": 1.2119,
+      "step": 8072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0420650761231527,
+      "learning_rate": 3.8554591056286336e-05,
+      "loss": 1.1914,
+      "step": 8073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2078162102694083,
+      "learning_rate": 3.855108652043845e-05,
+      "loss": 1.252,
+      "step": 8074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.630352015595886,
+      "learning_rate": 3.854758160746933e-05,
+      "loss": 1.2754,
+      "step": 8075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.176033491789816,
+      "learning_rate": 3.854407631747654e-05,
+      "loss": 1.4746,
+      "step": 8076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3983086923326713,
+      "learning_rate": 3.8540570650557603e-05,
+      "loss": 1.1328,
+      "step": 8077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.801945991234635,
+      "learning_rate": 3.8537064606810116e-05,
+      "loss": 1.252,
+      "step": 8078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.954864151312199,
+      "learning_rate": 3.853355818633162e-05,
+      "loss": 1.332,
+      "step": 8079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.06244035114385,
+      "learning_rate": 3.8530051389219715e-05,
+      "loss": 0.9751,
+      "step": 8080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.254255443213556,
+      "learning_rate": 3.852654421557198e-05,
+      "loss": 0.957,
+      "step": 8081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9830688022329555,
+      "learning_rate": 3.8523036665486026e-05,
+      "loss": 0.3358,
+      "step": 8082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.341385524609675,
+      "learning_rate": 3.8519528739059475e-05,
+      "loss": 0.9678,
+      "step": 8083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1367514431699624,
+      "learning_rate": 3.851602043638994e-05,
+      "loss": 0.8356,
+      "step": 8084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.488974917688192,
+      "learning_rate": 3.8512511757575065e-05,
+      "loss": 1.502,
+      "step": 8085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.190657924780156,
+      "learning_rate": 3.850900270271248e-05,
+      "loss": 0.6379,
+      "step": 8086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8274817815407878,
+      "learning_rate": 3.850549327189985e-05,
+      "loss": 0.8789,
+      "step": 8087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.662827524744261,
+      "learning_rate": 3.8501983465234844e-05,
+      "loss": 1.5781,
+      "step": 8088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.938245875855694,
+      "learning_rate": 3.849847328281513e-05,
+      "loss": 1.2246,
+      "step": 8089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.745019208670225,
+      "learning_rate": 3.8494962724738404e-05,
+      "loss": 1.3672,
+      "step": 8090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.091745242381849,
+      "learning_rate": 3.8491451791102354e-05,
+      "loss": 1.0312,
+      "step": 8091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7557547498853578,
+      "learning_rate": 3.8487940482004706e-05,
+      "loss": 0.761,
+      "step": 8092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0844229929242495,
+      "learning_rate": 3.848442879754315e-05,
+      "loss": 1.375,
+      "step": 8093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3969527131255552,
+      "learning_rate": 3.8480916737815445e-05,
+      "loss": 0.7402,
+      "step": 8094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.962753792164003,
+      "learning_rate": 3.847740430291929e-05,
+      "loss": 0.8884,
+      "step": 8095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8912342975881087,
+      "learning_rate": 3.847389149295249e-05,
+      "loss": 0.4619,
+      "step": 8096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8117862219530645,
+      "learning_rate": 3.8470378308012754e-05,
+      "loss": 1.0645,
+      "step": 8097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.228715357743897,
+      "learning_rate": 3.846686474819788e-05,
+      "loss": 0.687,
+      "step": 8098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2043589088990565,
+      "learning_rate": 3.846335081360565e-05,
+      "loss": 1.4219,
+      "step": 8099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2855485106706674,
+      "learning_rate": 3.845983650433384e-05,
+      "loss": 0.9639,
+      "step": 8100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4350536492486645,
+      "learning_rate": 3.845632182048026e-05,
+      "loss": 1.1426,
+      "step": 8101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5347869174302065,
+      "learning_rate": 3.8452806762142724e-05,
+      "loss": 0.7537,
+      "step": 8102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4520483873891847,
+      "learning_rate": 3.844929132941905e-05,
+      "loss": 0.8647,
+      "step": 8103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8418428935089732,
+      "learning_rate": 3.844577552240707e-05,
+      "loss": 0.6958,
+      "step": 8104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.993138095064765,
+      "learning_rate": 3.8442259341204635e-05,
+      "loss": 0.7026,
+      "step": 8105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.757201415199652,
+      "learning_rate": 3.8438742785909596e-05,
+      "loss": 1.0571,
+      "step": 8106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1432447751765697,
+      "learning_rate": 3.843522585661982e-05,
+      "loss": 1.0859,
+      "step": 8107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8029067993292816,
+      "learning_rate": 3.843170855343317e-05,
+      "loss": 1.6309,
+      "step": 8108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.798498439072062,
+      "learning_rate": 3.842819087644755e-05,
+      "loss": 1.4629,
+      "step": 8109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9750488275214235,
+      "learning_rate": 3.842467282576083e-05,
+      "loss": 1.1406,
+      "step": 8110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3935795895149954,
+      "learning_rate": 3.8421154401470935e-05,
+      "loss": 0.8633,
+      "step": 8111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4788543012858124,
+      "learning_rate": 3.8417635603675786e-05,
+      "loss": 0.9463,
+      "step": 8112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.262840508437913,
+      "learning_rate": 3.841411643247329e-05,
+      "loss": 1.2539,
+      "step": 8113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.987459095569654,
+      "learning_rate": 3.841059688796139e-05,
+      "loss": 0.7705,
+      "step": 8114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.677337634640802,
+      "learning_rate": 3.840707697023805e-05,
+      "loss": 1.7773,
+      "step": 8115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8321434883378207,
+      "learning_rate": 3.840355667940121e-05,
+      "loss": 1.2949,
+      "step": 8116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.696407845951,
+      "learning_rate": 3.840003601554885e-05,
+      "loss": 1.1699,
+      "step": 8117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.59912786790371,
+      "learning_rate": 3.839651497877894e-05,
+      "loss": 1.6045,
+      "step": 8118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.425340126022943,
+      "learning_rate": 3.8392993569189475e-05,
+      "loss": 0.7158,
+      "step": 8119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.19926212267417,
+      "learning_rate": 3.8389471786878446e-05,
+      "loss": 1.1094,
+      "step": 8120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.018206408624322,
+      "learning_rate": 3.838594963194387e-05,
+      "loss": 1.498,
+      "step": 8121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.398779856123237,
+      "learning_rate": 3.8382427104483765e-05,
+      "loss": 1.2324,
+      "step": 8122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.772847435805821,
+      "learning_rate": 3.837890420459616e-05,
+      "loss": 0.6758,
+      "step": 8123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.26796583758723,
+      "learning_rate": 3.83753809323791e-05,
+      "loss": 1.1758,
+      "step": 8124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.992639759291383,
+      "learning_rate": 3.837185728793063e-05,
+      "loss": 1.7754,
+      "step": 8125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5074731710107336,
+      "learning_rate": 3.836833327134882e-05,
+      "loss": 1.418,
+      "step": 8126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4662689408762555,
+      "learning_rate": 3.836480888273173e-05,
+      "loss": 0.8896,
+      "step": 8127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4777907926657647,
+      "learning_rate": 3.836128412217745e-05,
+      "loss": 1.2461,
+      "step": 8128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.438298901436278,
+      "learning_rate": 3.835775898978408e-05,
+      "loss": 0.9795,
+      "step": 8129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8859010996108783,
+      "learning_rate": 3.835423348564972e-05,
+      "loss": 0.572,
+      "step": 8130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8530741769855752,
+      "learning_rate": 3.835070760987247e-05,
+      "loss": 1.0254,
+      "step": 8131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.038832283468733,
+      "learning_rate": 3.834718136255046e-05,
+      "loss": 0.8047,
+      "step": 8132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.2483063501138485,
+      "learning_rate": 3.834365474378183e-05,
+      "loss": 1.793,
+      "step": 8133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.756268639187924,
+      "learning_rate": 3.8340127753664725e-05,
+      "loss": 0.7695,
+      "step": 8134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9299200754038317,
+      "learning_rate": 3.833660039229729e-05,
+      "loss": 0.9878,
+      "step": 8135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.24198559769145,
+      "learning_rate": 3.833307265977769e-05,
+      "loss": 1.3042,
+      "step": 8136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.1416736058216665,
+      "learning_rate": 3.832954455620411e-05,
+      "loss": 1.2217,
+      "step": 8137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7458699163644322,
+      "learning_rate": 3.8326016081674734e-05,
+      "loss": 0.9229,
+      "step": 8138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.994033708112938,
+      "learning_rate": 3.8322487236287764e-05,
+      "loss": 0.7812,
+      "step": 8139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.644517287352856,
+      "learning_rate": 3.831895802014138e-05,
+      "loss": 1.1621,
+      "step": 8140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9216292710851115,
+      "learning_rate": 3.8315428433333835e-05,
+      "loss": 0.6113,
+      "step": 8141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.142102401696447,
+      "learning_rate": 3.831189847596331e-05,
+      "loss": 0.7249,
+      "step": 8142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.218127367716124,
+      "learning_rate": 3.83083681481281e-05,
+      "loss": 1.209,
+      "step": 8143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.84534088658415,
+      "learning_rate": 3.8304837449926406e-05,
+      "loss": 1.2576,
+      "step": 8144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6322710063341415,
+      "learning_rate": 3.8301306381456506e-05,
+      "loss": 1.0918,
+      "step": 8145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7507501487658166,
+      "learning_rate": 3.8297774942816666e-05,
+      "loss": 0.8027,
+      "step": 8146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.7797856306298385,
+      "learning_rate": 3.829424313410516e-05,
+      "loss": 0.9668,
+      "step": 8147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.641954636101901,
+      "learning_rate": 3.8290710955420276e-05,
+      "loss": 1.3164,
+      "step": 8148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.256542555379522,
+      "learning_rate": 3.8287178406860324e-05,
+      "loss": 0.8486,
+      "step": 8149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1602875218988125,
+      "learning_rate": 3.82836454885236e-05,
+      "loss": 1.123,
+      "step": 8150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.391775415409304,
+      "learning_rate": 3.828011220050843e-05,
+      "loss": 0.3883,
+      "step": 8151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.923952088100448,
+      "learning_rate": 3.827657854291315e-05,
+      "loss": 1.3555,
+      "step": 8152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3368322387602785,
+      "learning_rate": 3.82730445158361e-05,
+      "loss": 0.8743,
+      "step": 8153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.370662040310448,
+      "learning_rate": 3.826951011937561e-05,
+      "loss": 0.9043,
+      "step": 8154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.990540035341128,
+      "learning_rate": 3.826597535363006e-05,
+      "loss": 1.6445,
+      "step": 8155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.183143716923466,
+      "learning_rate": 3.826244021869782e-05,
+      "loss": 1.2129,
+      "step": 8156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.093008738715875,
+      "learning_rate": 3.8258904714677264e-05,
+      "loss": 1.2168,
+      "step": 8157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.086970103093746,
+      "learning_rate": 3.8255368841666796e-05,
+      "loss": 1.1895,
+      "step": 8158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9082063110699796,
+      "learning_rate": 3.825183259976479e-05,
+      "loss": 1.3594,
+      "step": 8159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.508889121396285,
+      "learning_rate": 3.82482959890697e-05,
+      "loss": 2.0137,
+      "step": 8160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.746294885462818,
+      "learning_rate": 3.824475900967991e-05,
+      "loss": 0.5787,
+      "step": 8161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.574678556507985,
+      "learning_rate": 3.824122166169388e-05,
+      "loss": 1.4512,
+      "step": 8162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8277885279006654,
+      "learning_rate": 3.823768394521004e-05,
+      "loss": 0.9346,
+      "step": 8163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.374926672244863,
+      "learning_rate": 3.8234145860326835e-05,
+      "loss": 1.6289,
+      "step": 8164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2930470441874258,
+      "learning_rate": 3.8230607407142736e-05,
+      "loss": 1.0215,
+      "step": 8165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.721209542954382,
+      "learning_rate": 3.8227068585756234e-05,
+      "loss": 2.1953,
+      "step": 8166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.653208708144651,
+      "learning_rate": 3.822352939626579e-05,
+      "loss": 0.6846,
+      "step": 8167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.760705228221921,
+      "learning_rate": 3.821998983876991e-05,
+      "loss": 0.8599,
+      "step": 8168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2242488150244557,
+      "learning_rate": 3.821644991336709e-05,
+      "loss": 1.1995,
+      "step": 8169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.184164284378302,
+      "learning_rate": 3.821290962015585e-05,
+      "loss": 1.0801,
+      "step": 8170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2119212843080227,
+      "learning_rate": 3.820936895923472e-05,
+      "loss": 1.0308,
+      "step": 8171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.095688901379153,
+      "learning_rate": 3.820582793070222e-05,
+      "loss": 1.1953,
+      "step": 8172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.995926352861926,
+      "learning_rate": 3.8202286534656914e-05,
+      "loss": 1.3594,
+      "step": 8173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.028033408087765,
+      "learning_rate": 3.819874477119734e-05,
+      "loss": 1.2402,
+      "step": 8174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.106150010748098,
+      "learning_rate": 3.819520264042208e-05,
+      "loss": 1.0664,
+      "step": 8175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.323122115237313,
+      "learning_rate": 3.8191660142429694e-05,
+      "loss": 0.8135,
+      "step": 8176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.194378796892489,
+      "learning_rate": 3.818811727731879e-05,
+      "loss": 1.3496,
+      "step": 8177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.09919895152845,
+      "learning_rate": 3.8184574045187936e-05,
+      "loss": 0.9648,
+      "step": 8178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.609213178722241,
+      "learning_rate": 3.8181030446135765e-05,
+      "loss": 1.0708,
+      "step": 8179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8024026814015874,
+      "learning_rate": 3.817748648026087e-05,
+      "loss": 1.0205,
+      "step": 8180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.465853550768823,
+      "learning_rate": 3.8173942147661904e-05,
+      "loss": 1.1045,
+      "step": 8181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.188877122901202,
+      "learning_rate": 3.817039744843749e-05,
+      "loss": 0.9824,
+      "step": 8182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9880827395257903,
+      "learning_rate": 3.816685238268627e-05,
+      "loss": 1.2197,
+      "step": 8183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.50081394674658,
+      "learning_rate": 3.81633069505069e-05,
+      "loss": 0.4546,
+      "step": 8184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.259512258914764,
+      "learning_rate": 3.815976115199808e-05,
+      "loss": 0.584,
+      "step": 8185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1985393443644035,
+      "learning_rate": 3.815621498725845e-05,
+      "loss": 1.3398,
+      "step": 8186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.459596012007186,
+      "learning_rate": 3.815266845638672e-05,
+      "loss": 1.5859,
+      "step": 8187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7110808496749264,
+      "learning_rate": 3.814912155948157e-05,
+      "loss": 0.9258,
+      "step": 8188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.653322558916283,
+      "learning_rate": 3.814557429664173e-05,
+      "loss": 0.8613,
+      "step": 8189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.785276177647866,
+      "learning_rate": 3.81420266679659e-05,
+      "loss": 1.0117,
+      "step": 8190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.672553277934007,
+      "learning_rate": 3.813847867355282e-05,
+      "loss": 0.9434,
+      "step": 8191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.208237725610213,
+      "learning_rate": 3.813493031350124e-05,
+      "loss": 0.9634,
+      "step": 8192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5048810992318704,
+      "learning_rate": 3.8131381587909875e-05,
+      "loss": 0.7991,
+      "step": 8193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.112268857634518,
+      "learning_rate": 3.812783249687752e-05,
+      "loss": 0.6836,
+      "step": 8194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.47868313351242,
+      "learning_rate": 3.812428304050293e-05,
+      "loss": 0.9482,
+      "step": 8195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6894956764063043,
+      "learning_rate": 3.8120733218884894e-05,
+      "loss": 1.2578,
+      "step": 8196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2340343722548264,
+      "learning_rate": 3.811718303212218e-05,
+      "loss": 1.5527,
+      "step": 8197
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4966929497075987,
+      "learning_rate": 3.8113632480313605e-05,
+      "loss": 1.5762,
+      "step": 8198
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3184745522147914,
+      "learning_rate": 3.811008156355798e-05,
+      "loss": 0.9199,
+      "step": 8199
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.063944417573846,
+      "learning_rate": 3.810653028195412e-05,
+      "loss": 1.4395,
+      "step": 8200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4411630738267234,
+      "learning_rate": 3.810297863560086e-05,
+      "loss": 0.8896,
+      "step": 8201
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.818830995576898,
+      "learning_rate": 3.8099426624597034e-05,
+      "loss": 1.0645,
+      "step": 8202
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.187290354449314,
+      "learning_rate": 3.8095874249041494e-05,
+      "loss": 1.2378,
+      "step": 8203
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.546015498659578,
+      "learning_rate": 3.809232150903312e-05,
+      "loss": 1.877,
+      "step": 8204
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.504140856042828,
+      "learning_rate": 3.808876840467076e-05,
+      "loss": 0.8672,
+      "step": 8205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.555435982869099,
+      "learning_rate": 3.80852149360533e-05,
+      "loss": 1.0332,
+      "step": 8206
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7001469866450334,
+      "learning_rate": 3.808166110327964e-05,
+      "loss": 1.5684,
+      "step": 8207
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.581246426299278,
+      "learning_rate": 3.8078106906448683e-05,
+      "loss": 1.0078,
+      "step": 8208
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3835458096134587,
+      "learning_rate": 3.807455234565933e-05,
+      "loss": 0.7979,
+      "step": 8209
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.459371595368216,
+      "learning_rate": 3.80709974210105e-05,
+      "loss": 0.7373,
+      "step": 8210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.741487122162544,
+      "learning_rate": 3.806744213260115e-05,
+      "loss": 1.9678,
+      "step": 8211
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2305266111051605,
+      "learning_rate": 3.806388648053019e-05,
+      "loss": 1.501,
+      "step": 8212
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8044138485330574,
+      "learning_rate": 3.806033046489659e-05,
+      "loss": 0.7275,
+      "step": 8213
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.350535763447677,
+      "learning_rate": 3.8056774085799315e-05,
+      "loss": 1.2539,
+      "step": 8214
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.3799902657178595,
+      "learning_rate": 3.8053217343337334e-05,
+      "loss": 1.4668,
+      "step": 8215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.295056823658626,
+      "learning_rate": 3.8049660237609614e-05,
+      "loss": 1.1992,
+      "step": 8216
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8963737416237132,
+      "learning_rate": 3.804610276871518e-05,
+      "loss": 0.7183,
+      "step": 8217
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.23980274929215,
+      "learning_rate": 3.8042544936753e-05,
+      "loss": 0.7778,
+      "step": 8218
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4344670916119613,
+      "learning_rate": 3.803898674182212e-05,
+      "loss": 0.8926,
+      "step": 8219
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.134169214899494,
+      "learning_rate": 3.8035428184021535e-05,
+      "loss": 0.9851,
+      "step": 8220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.493289782080341,
+      "learning_rate": 3.80318692634503e-05,
+      "loss": 0.7041,
+      "step": 8221
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.767148304192647,
+      "learning_rate": 3.8028309980207446e-05,
+      "loss": 0.9463,
+      "step": 8222
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6644546354429637,
+      "learning_rate": 3.8024750334392025e-05,
+      "loss": 1.0674,
+      "step": 8223
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0748375151108407,
+      "learning_rate": 3.8021190326103104e-05,
+      "loss": 1.2168,
+      "step": 8224
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.990305466714743,
+      "learning_rate": 3.801762995543976e-05,
+      "loss": 1.2031,
+      "step": 8225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7956181710513293,
+      "learning_rate": 3.801406922250108e-05,
+      "loss": 0.6885,
+      "step": 8226
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.108480958175589,
+      "learning_rate": 3.801050812738614e-05,
+      "loss": 1.2549,
+      "step": 8227
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.122009793054075,
+      "learning_rate": 3.800694667019407e-05,
+      "loss": 1.8535,
+      "step": 8228
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5019853955306917,
+      "learning_rate": 3.8003384851023955e-05,
+      "loss": 1.0186,
+      "step": 8229
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.549931589340747,
+      "learning_rate": 3.7999822669974935e-05,
+      "loss": 0.8877,
+      "step": 8230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.325155630337838,
+      "learning_rate": 3.799626012714614e-05,
+      "loss": 0.834,
+      "step": 8231
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.943173876967515,
+      "learning_rate": 3.799269722263673e-05,
+      "loss": 1.093,
+      "step": 8232
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.237780907570787,
+      "learning_rate": 3.798913395654584e-05,
+      "loss": 1.3633,
+      "step": 8233
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.463055380529669,
+      "learning_rate": 3.7985570328972644e-05,
+      "loss": 1.3047,
+      "step": 8234
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.988771086793522,
+      "learning_rate": 3.79820063400163e-05,
+      "loss": 0.8711,
+      "step": 8235
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.85507257000698,
+      "learning_rate": 3.797844198977601e-05,
+      "loss": 1.853,
+      "step": 8236
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.799973869252604,
+      "learning_rate": 3.797487727835096e-05,
+      "loss": 1.7051,
+      "step": 8237
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.765803036101844,
+      "learning_rate": 3.797131220584036e-05,
+      "loss": 0.8232,
+      "step": 8238
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.388860529839179,
+      "learning_rate": 3.796774677234342e-05,
+      "loss": 1.168,
+      "step": 8239
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.391468634449697,
+      "learning_rate": 3.7964180977959376e-05,
+      "loss": 0.959,
+      "step": 8240
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.19308983834769,
+      "learning_rate": 3.796061482278745e-05,
+      "loss": 0.9097,
+      "step": 8241
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9168633891813585,
+      "learning_rate": 3.795704830692688e-05,
+      "loss": 1.8047,
+      "step": 8242
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3904288217670935,
+      "learning_rate": 3.795348143047693e-05,
+      "loss": 1.2012,
+      "step": 8243
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9830341622432606,
+      "learning_rate": 3.7949914193536877e-05,
+      "loss": 1.1152,
+      "step": 8244
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2260406293775996,
+      "learning_rate": 3.7946346596205985e-05,
+      "loss": 0.8984,
+      "step": 8245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.641966757442072,
+      "learning_rate": 3.794277863858353e-05,
+      "loss": 0.9722,
+      "step": 8246
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.111640334790457,
+      "learning_rate": 3.7939210320768814e-05,
+      "loss": 1.0723,
+      "step": 8247
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.790786193791601,
+      "learning_rate": 3.793564164286114e-05,
+      "loss": 0.9463,
+      "step": 8248
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1564176190227706,
+      "learning_rate": 3.7932072604959834e-05,
+      "loss": 0.1597,
+      "step": 8249
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.149553979614627,
+      "learning_rate": 3.7928503207164204e-05,
+      "loss": 1.2383,
+      "step": 8250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.306325278987558,
+      "learning_rate": 3.79249334495736e-05,
+      "loss": 0.7803,
+      "step": 8251
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0061271404452414,
+      "learning_rate": 3.792136333228735e-05,
+      "loss": 0.7065,
+      "step": 8252
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.733646147872851,
+      "learning_rate": 3.7917792855404834e-05,
+      "loss": 1.1035,
+      "step": 8253
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.154527704160469,
+      "learning_rate": 3.791422201902539e-05,
+      "loss": 0.7012,
+      "step": 8254
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8981012417214678,
+      "learning_rate": 3.7910650823248404e-05,
+      "loss": 1.5449,
+      "step": 8255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.044405361800233,
+      "learning_rate": 3.7907079268173265e-05,
+      "loss": 0.9917,
+      "step": 8256
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8153527944321115,
+      "learning_rate": 3.790350735389937e-05,
+      "loss": 1.209,
+      "step": 8257
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4597893837188294,
+      "learning_rate": 3.789993508052611e-05,
+      "loss": 1.2285,
+      "step": 8258
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.62993430545114,
+      "learning_rate": 3.789636244815291e-05,
+      "loss": 1.0522,
+      "step": 8259
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.928736304254112,
+      "learning_rate": 3.78927894568792e-05,
+      "loss": 1.2363,
+      "step": 8260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.0512474476431235,
+      "learning_rate": 3.788921610680439e-05,
+      "loss": 0.8154,
+      "step": 8261
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.583824746758272,
+      "learning_rate": 3.788564239802796e-05,
+      "loss": 0.2498,
+      "step": 8262
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.494828354272936,
+      "learning_rate": 3.788206833064935e-05,
+      "loss": 0.8438,
+      "step": 8263
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.4983854576352025,
+      "learning_rate": 3.787849390476801e-05,
+      "loss": 0.6953,
+      "step": 8264
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8746143272192115,
+      "learning_rate": 3.7874919120483434e-05,
+      "loss": 1.1523,
+      "step": 8265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.938858038614386,
+      "learning_rate": 3.78713439778951e-05,
+      "loss": 1.4121,
+      "step": 8266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6175154667057705,
+      "learning_rate": 3.786776847710249e-05,
+      "loss": 1.2031,
+      "step": 8267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6505737691190547,
+      "learning_rate": 3.786419261820514e-05,
+      "loss": 1.2871,
+      "step": 8268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0139016084261008,
+      "learning_rate": 3.786061640130254e-05,
+      "loss": 1.1104,
+      "step": 8269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4865075671607455,
+      "learning_rate": 3.785703982649423e-05,
+      "loss": 0.9434,
+      "step": 8270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.384026490185921,
+      "learning_rate": 3.785346289387972e-05,
+      "loss": 0.9473,
+      "step": 8271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.727941189033858,
+      "learning_rate": 3.7849885603558575e-05,
+      "loss": 0.9941,
+      "step": 8272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.004393364024708,
+      "learning_rate": 3.784630795563035e-05,
+      "loss": 1.3818,
+      "step": 8273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.679025561603479,
+      "learning_rate": 3.78427299501946e-05,
+      "loss": 1.1201,
+      "step": 8274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.584098138705302,
+      "learning_rate": 3.783915158735091e-05,
+      "loss": 1.1992,
+      "step": 8275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.451888491636754,
+      "learning_rate": 3.7835572867198854e-05,
+      "loss": 1.3672,
+      "step": 8276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.366176779733078,
+      "learning_rate": 3.783199378983803e-05,
+      "loss": 1.0146,
+      "step": 8277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.635815813388006,
+      "learning_rate": 3.782841435536804e-05,
+      "loss": 1.2708,
+      "step": 8278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.025514662916857,
+      "learning_rate": 3.782483456388851e-05,
+      "loss": 0.9326,
+      "step": 8279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.668427680605396,
+      "learning_rate": 3.782125441549905e-05,
+      "loss": 1.3398,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.812792570498722,
+      "learning_rate": 3.781767391029929e-05,
+      "loss": 1.875,
+      "step": 8281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.212947763787128,
+      "learning_rate": 3.7814093048388894e-05,
+      "loss": 1.002,
+      "step": 8282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.006268334230685,
+      "learning_rate": 3.78105118298675e-05,
+      "loss": 1.04,
+      "step": 8283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7984359051110403,
+      "learning_rate": 3.780693025483478e-05,
+      "loss": 1.2031,
+      "step": 8284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.567436996050479,
+      "learning_rate": 3.7803348323390405e-05,
+      "loss": 1.2559,
+      "step": 8285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.504047692651109,
+      "learning_rate": 3.7799766035634057e-05,
+      "loss": 0.8711,
+      "step": 8286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.741784238603518,
+      "learning_rate": 3.779618339166543e-05,
+      "loss": 0.8984,
+      "step": 8287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9164871887208172,
+      "learning_rate": 3.7792600391584234e-05,
+      "loss": 0.8495,
+      "step": 8288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.481404776907308,
+      "learning_rate": 3.778901703549017e-05,
+      "loss": 1.3418,
+      "step": 8289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3647318867267364,
+      "learning_rate": 3.778543332348297e-05,
+      "loss": 1.4824,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.106609705892598,
+      "learning_rate": 3.778184925566237e-05,
+      "loss": 0.6399,
+      "step": 8291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.02516066069395,
+      "learning_rate": 3.77782648321281e-05,
+      "loss": 0.9082,
+      "step": 8292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.736661467305246,
+      "learning_rate": 3.7774680052979934e-05,
+      "loss": 0.8711,
+      "step": 8293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8854852147059744,
+      "learning_rate": 3.777109491831761e-05,
+      "loss": 1.0254,
+      "step": 8294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3444046377922962,
+      "learning_rate": 3.7767509428240924e-05,
+      "loss": 1.4355,
+      "step": 8295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.844244777688396,
+      "learning_rate": 3.776392358284965e-05,
+      "loss": 1.1709,
+      "step": 8296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.12006478264632,
+      "learning_rate": 3.7760337382243566e-05,
+      "loss": 0.4854,
+      "step": 8297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.1048691563075765,
+      "learning_rate": 3.77567508265225e-05,
+      "loss": 0.9111,
+      "step": 8298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9396662739372346,
+      "learning_rate": 3.775316391578624e-05,
+      "loss": 0.6626,
+      "step": 8299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1720784044042127,
+      "learning_rate": 3.774957665013462e-05,
+      "loss": 1.168,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.748134403290116,
+      "learning_rate": 3.774598902966748e-05,
+      "loss": 1.2363,
+      "step": 8301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.799565645996517,
+      "learning_rate": 3.774240105448466e-05,
+      "loss": 0.8577,
+      "step": 8302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9453520215764923,
+      "learning_rate": 3.773881272468599e-05,
+      "loss": 0.7617,
+      "step": 8303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.111922470692709,
+      "learning_rate": 3.7735224040371355e-05,
+      "loss": 1.0957,
+      "step": 8304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1006837730732166,
+      "learning_rate": 3.773163500164062e-05,
+      "loss": 0.9639,
+      "step": 8305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.478710602770888,
+      "learning_rate": 3.772804560859367e-05,
+      "loss": 1.5293,
+      "step": 8306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.4327510396900145,
+      "learning_rate": 3.772445586133039e-05,
+      "loss": 1.1211,
+      "step": 8307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2701741816371546,
+      "learning_rate": 3.772086575995069e-05,
+      "loss": 1.3242,
+      "step": 8308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.464899511183238,
+      "learning_rate": 3.7717275304554454e-05,
+      "loss": 1.2588,
+      "step": 8309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.00313851082459,
+      "learning_rate": 3.771368449524164e-05,
+      "loss": 0.7549,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4082725189926792,
+      "learning_rate": 3.771009333211216e-05,
+      "loss": 0.8848,
+      "step": 8311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.478269880550621,
+      "learning_rate": 3.770650181526596e-05,
+      "loss": 1.1466,
+      "step": 8312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0372475350407857,
+      "learning_rate": 3.770290994480298e-05,
+      "loss": 0.9248,
+      "step": 8313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.10125051719196,
+      "learning_rate": 3.769931772082319e-05,
+      "loss": 1.6494,
+      "step": 8314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4950685500090746,
+      "learning_rate": 3.769572514342656e-05,
+      "loss": 0.7017,
+      "step": 8315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0990107476921125,
+      "learning_rate": 3.769213221271306e-05,
+      "loss": 1.5039,
+      "step": 8316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4619929854808422,
+      "learning_rate": 3.76885389287827e-05,
+      "loss": 1.1406,
+      "step": 8317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.908876142249566,
+      "learning_rate": 3.768494529173545e-05,
+      "loss": 0.8682,
+      "step": 8318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.543534035852583,
+      "learning_rate": 3.7681351301671354e-05,
+      "loss": 1.1953,
+      "step": 8319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.040198136673451,
+      "learning_rate": 3.76777569586904e-05,
+      "loss": 0.918,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.750667324371963,
+      "learning_rate": 3.767416226289264e-05,
+      "loss": 1.4922,
+      "step": 8321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.464810001649575,
+      "learning_rate": 3.7670567214378104e-05,
+      "loss": 1.4199,
+      "step": 8322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1953335413904544,
+      "learning_rate": 3.766697181324684e-05,
+      "loss": 1.377,
+      "step": 8323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.604906378910652,
+      "learning_rate": 3.7663376059598895e-05,
+      "loss": 0.8726,
+      "step": 8324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.147558316676738,
+      "learning_rate": 3.765977995353437e-05,
+      "loss": 0.5005,
+      "step": 8325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3664438823378835,
+      "learning_rate": 3.765618349515331e-05,
+      "loss": 0.2649,
+      "step": 8326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.528889850391323,
+      "learning_rate": 3.765258668455582e-05,
+      "loss": 1.2441,
+      "step": 8327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3492474863729593,
+      "learning_rate": 3.764898952184199e-05,
+      "loss": 0.9111,
+      "step": 8328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.054943344374751,
+      "learning_rate": 3.7645392007111934e-05,
+      "loss": 0.7568,
+      "step": 8329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.556332682243174,
+      "learning_rate": 3.764179414046577e-05,
+      "loss": 1.5002,
+      "step": 8330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5094425347777425,
+      "learning_rate": 3.763819592200362e-05,
+      "loss": 1.3037,
+      "step": 8331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9981960267063439,
+      "learning_rate": 3.763459735182561e-05,
+      "loss": 0.4766,
+      "step": 8332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6357501476121,
+      "learning_rate": 3.7630998430031906e-05,
+      "loss": 1.3359,
+      "step": 8333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.681569938778529,
+      "learning_rate": 3.7627399156722666e-05,
+      "loss": 1.2227,
+      "step": 8334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.407307082990283,
+      "learning_rate": 3.762379953199804e-05,
+      "loss": 1.127,
+      "step": 8335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.371447063938596,
+      "learning_rate": 3.762019955595822e-05,
+      "loss": 0.895,
+      "step": 8336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.569148406101778,
+      "learning_rate": 3.761659922870337e-05,
+      "loss": 0.957,
+      "step": 8337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.456005707259234,
+      "learning_rate": 3.761299855033371e-05,
+      "loss": 1.084,
+      "step": 8338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.718702972719036,
+      "learning_rate": 3.760939752094943e-05,
+      "loss": 0.811,
+      "step": 8339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.2861711496104276,
+      "learning_rate": 3.7605796140650764e-05,
+      "loss": 1.627,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.691209406637383,
+      "learning_rate": 3.76021944095379e-05,
+      "loss": 1.2207,
+      "step": 8341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3404171452842397,
+      "learning_rate": 3.759859232771111e-05,
+      "loss": 0.5239,
+      "step": 8342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7049628388113924,
+      "learning_rate": 3.7594989895270625e-05,
+      "loss": 0.9512,
+      "step": 8343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8069760687085656,
+      "learning_rate": 3.75913871123167e-05,
+      "loss": 0.8271,
+      "step": 8344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.807124192450708,
+      "learning_rate": 3.75877839789496e-05,
+      "loss": 1.1738,
+      "step": 8345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.377814341380042,
+      "learning_rate": 3.758418049526958e-05,
+      "loss": 1.2236,
+      "step": 8346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.397615273513771,
+      "learning_rate": 3.758057666137695e-05,
+      "loss": 1.2461,
+      "step": 8347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.956848862631648,
+      "learning_rate": 3.7576972477371994e-05,
+      "loss": 1.2891,
+      "step": 8348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.523243321811842,
+      "learning_rate": 3.757336794335502e-05,
+      "loss": 0.6323,
+      "step": 8349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5636206246722373,
+      "learning_rate": 3.7569763059426333e-05,
+      "loss": 1.125,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.287787164054116,
+      "learning_rate": 3.756615782568625e-05,
+      "loss": 2.7168,
+      "step": 8351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.583569341315878,
+      "learning_rate": 3.756255224223511e-05,
+      "loss": 2.291,
+      "step": 8352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.8398411371231616,
+      "learning_rate": 3.755894630917326e-05,
+      "loss": 1.375,
+      "step": 8353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.478035124159924,
+      "learning_rate": 3.755534002660105e-05,
+      "loss": 0.7617,
+      "step": 8354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8651669734211818,
+      "learning_rate": 3.7551733394618825e-05,
+      "loss": 1.1875,
+      "step": 8355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.96131057871656,
+      "learning_rate": 3.754812641332698e-05,
+      "loss": 1.0498,
+      "step": 8356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.091550046627515,
+      "learning_rate": 3.754451908282589e-05,
+      "loss": 1.1929,
+      "step": 8357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8910510000959755,
+      "learning_rate": 3.754091140321593e-05,
+      "loss": 0.5455,
+      "step": 8358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7354687481618223,
+      "learning_rate": 3.7537303374597514e-05,
+      "loss": 0.8735,
+      "step": 8359
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.57273708648599,
+      "learning_rate": 3.7533694997071044e-05,
+      "loss": 1.2588,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5066831904062847,
+      "learning_rate": 3.753008627073696e-05,
+      "loss": 1.4268,
+      "step": 8361
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.772625785567835,
+      "learning_rate": 3.7526477195695664e-05,
+      "loss": 1.4941,
+      "step": 8362
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.411563001692693,
+      "learning_rate": 3.752286777204761e-05,
+      "loss": 1.7754,
+      "step": 8363
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.682654994748095,
+      "learning_rate": 3.7519257999893244e-05,
+      "loss": 1.6133,
+      "step": 8364
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.964751914390038,
+      "learning_rate": 3.751564787933303e-05,
+      "loss": 0.9844,
+      "step": 8365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.303059292515308,
+      "learning_rate": 3.751203741046742e-05,
+      "loss": 0.8843,
+      "step": 8366
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.888645554588931,
+      "learning_rate": 3.7508426593396906e-05,
+      "loss": 0.6055,
+      "step": 8367
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.441983299068561,
+      "learning_rate": 3.750481542822198e-05,
+      "loss": 1.5664,
+      "step": 8368
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4950649346034655,
+      "learning_rate": 3.7501203915043114e-05,
+      "loss": 1.3486,
+      "step": 8369
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.158983280646511,
+      "learning_rate": 3.7497592053960854e-05,
+      "loss": 1.1621,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8413005010305121,
+      "learning_rate": 3.749397984507568e-05,
+      "loss": 0.8353,
+      "step": 8371
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.404138208107354,
+      "learning_rate": 3.7490367288488145e-05,
+      "loss": 0.7876,
+      "step": 8372
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.123234520334158,
+      "learning_rate": 3.748675438429876e-05,
+      "loss": 0.7717,
+      "step": 8373
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6601148214204215,
+      "learning_rate": 3.748314113260809e-05,
+      "loss": 0.8662,
+      "step": 8374
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.000099656483211,
+      "learning_rate": 3.747952753351668e-05,
+      "loss": 0.2454,
+      "step": 8375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2076883873742394,
+      "learning_rate": 3.74759135871251e-05,
+      "loss": 1.0908,
+      "step": 8376
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8813790579996423,
+      "learning_rate": 3.747229929353393e-05,
+      "loss": 1.1973,
+      "step": 8377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.327781838648262,
+      "learning_rate": 3.7468684652843745e-05,
+      "loss": 1.3301,
+      "step": 8378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.71204855297054,
+      "learning_rate": 3.7465069665155136e-05,
+      "loss": 1.4492,
+      "step": 8379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9291903875122123,
+      "learning_rate": 3.746145433056873e-05,
+      "loss": 0.73,
+      "step": 8380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.563965850736252,
+      "learning_rate": 3.745783864918511e-05,
+      "loss": 0.9414,
+      "step": 8381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4252682449005305,
+      "learning_rate": 3.745422262110492e-05,
+      "loss": 0.875,
+      "step": 8382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.800717391509728,
+      "learning_rate": 3.745060624642877e-05,
+      "loss": 1.0361,
+      "step": 8383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6371629580971696,
+      "learning_rate": 3.744698952525733e-05,
+      "loss": 1.1387,
+      "step": 8384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2974074388855326,
+      "learning_rate": 3.7443372457691236e-05,
+      "loss": 1.2051,
+      "step": 8385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3731972729154007,
+      "learning_rate": 3.743975504383115e-05,
+      "loss": 0.5894,
+      "step": 8386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.037784495472618,
+      "learning_rate": 3.743613728377774e-05,
+      "loss": 1.0737,
+      "step": 8387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1861741543553963,
+      "learning_rate": 3.7432519177631696e-05,
+      "loss": 0.531,
+      "step": 8388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.023600413741058,
+      "learning_rate": 3.742890072549371e-05,
+      "loss": 0.9102,
+      "step": 8389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.156677535421272,
+      "learning_rate": 3.7425281927464476e-05,
+      "loss": 0.8574,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.56189698797617,
+      "learning_rate": 3.74216627836447e-05,
+      "loss": 1.3047,
+      "step": 8391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8702065137227892,
+      "learning_rate": 3.7418043294135095e-05,
+      "loss": 1.0308,
+      "step": 8392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.910284216198682,
+      "learning_rate": 3.7414423459036416e-05,
+      "loss": 1.3594,
+      "step": 8393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.233875581060603,
+      "learning_rate": 3.741080327844939e-05,
+      "loss": 1.0361,
+      "step": 8394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.728530926191575,
+      "learning_rate": 3.740718275247475e-05,
+      "loss": 0.8564,
+      "step": 8395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.92894628706487,
+      "learning_rate": 3.740356188121327e-05,
+      "loss": 1.3242,
+      "step": 8396
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.220747502061362,
+      "learning_rate": 3.739994066476571e-05,
+      "loss": 1.6719,
+      "step": 8397
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.920409517695567,
+      "learning_rate": 3.739631910323285e-05,
+      "loss": 0.8633,
+      "step": 8398
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2214043012702906,
+      "learning_rate": 3.739269719671548e-05,
+      "loss": 1.2012,
+      "step": 8399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9146757428017,
+      "learning_rate": 3.738907494531438e-05,
+      "loss": 0.7236,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.395296606121845,
+      "learning_rate": 3.738545234913038e-05,
+      "loss": 1.293,
+      "step": 8401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9993294545477656,
+      "learning_rate": 3.738182940826428e-05,
+      "loss": 1.2402,
+      "step": 8402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.52816232119056,
+      "learning_rate": 3.737820612281691e-05,
+      "loss": 2.2393,
+      "step": 8403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.201954504915177,
+      "learning_rate": 3.73745824928891e-05,
+      "loss": 0.6543,
+      "step": 8404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8965082456977433,
+      "learning_rate": 3.7370958518581695e-05,
+      "loss": 1.0625,
+      "step": 8405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.759168517224549,
+      "learning_rate": 3.7367334199995555e-05,
+      "loss": 0.8076,
+      "step": 8406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3439109238373965,
+      "learning_rate": 3.736370953723153e-05,
+      "loss": 1.2715,
+      "step": 8407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.559161445567997,
+      "learning_rate": 3.736008453039051e-05,
+      "loss": 0.978,
+      "step": 8408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9051953114495515,
+      "learning_rate": 3.7356459179573365e-05,
+      "loss": 1.2129,
+      "step": 8409
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.539763368562559,
+      "learning_rate": 3.7352833484881e-05,
+      "loss": 0.688,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.110943745021506,
+      "learning_rate": 3.73492074464143e-05,
+      "loss": 0.8965,
+      "step": 8411
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5794804564488207,
+      "learning_rate": 3.734558106427418e-05,
+      "loss": 1.0938,
+      "step": 8412
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.234233078083974,
+      "learning_rate": 3.734195433856157e-05,
+      "loss": 0.4438,
+      "step": 8413
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5721196369983397,
+      "learning_rate": 3.7338327269377404e-05,
+      "loss": 0.665,
+      "step": 8414
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.872724418663507,
+      "learning_rate": 3.7334699856822606e-05,
+      "loss": 1.2012,
+      "step": 8415
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1594035191668426,
+      "learning_rate": 3.7331072100998124e-05,
+      "loss": 1.0566,
+      "step": 8416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.876670318383116,
+      "learning_rate": 3.732744400200494e-05,
+      "loss": 1.6484,
+      "step": 8417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0330153789923684,
+      "learning_rate": 3.7323815559943996e-05,
+      "loss": 0.6865,
+      "step": 8418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9873583197416718,
+      "learning_rate": 3.732018677491629e-05,
+      "loss": 1.6211,
+      "step": 8419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8069329159354286,
+      "learning_rate": 3.73165576470228e-05,
+      "loss": 0.791,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1829034741284508,
+      "learning_rate": 3.731292817636453e-05,
+      "loss": 0.8096,
+      "step": 8421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.693689353109022,
+      "learning_rate": 3.730929836304247e-05,
+      "loss": 1.4111,
+      "step": 8422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0768324563546745,
+      "learning_rate": 3.730566820715765e-05,
+      "loss": 0.9155,
+      "step": 8423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5176021790816554,
+      "learning_rate": 3.73020377088111e-05,
+      "loss": 0.8848,
+      "step": 8424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8867583655706046,
+      "learning_rate": 3.729840686810384e-05,
+      "loss": 0.5378,
+      "step": 8425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7743040638393,
+      "learning_rate": 3.729477568513693e-05,
+      "loss": 1.1133,
+      "step": 8426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.660916058938597,
+      "learning_rate": 3.7291144160011415e-05,
+      "loss": 2.4102,
+      "step": 8427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4008567179485265,
+      "learning_rate": 3.728751229282836e-05,
+      "loss": 0.917,
+      "step": 8428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.578676184046705,
+      "learning_rate": 3.728388008368884e-05,
+      "loss": 1.0908,
+      "step": 8429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6066577034460376,
+      "learning_rate": 3.728024753269394e-05,
+      "loss": 1.0762,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.19476583061588,
+      "learning_rate": 3.727661463994476e-05,
+      "loss": 0.915,
+      "step": 8431
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5950867299192955,
+      "learning_rate": 3.727298140554238e-05,
+      "loss": 1.4062,
+      "step": 8432
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.338210749579175,
+      "learning_rate": 3.726934782958793e-05,
+      "loss": 1.0771,
+      "step": 8433
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6387985235111213,
+      "learning_rate": 3.726571391218253e-05,
+      "loss": 0.3271,
+      "step": 8434
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6407934733261906,
+      "learning_rate": 3.726207965342729e-05,
+      "loss": 0.9609,
+      "step": 8435
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.547938586993576,
+      "learning_rate": 3.7258445053423376e-05,
+      "loss": 1.1514,
+      "step": 8436
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.912145830639997,
+      "learning_rate": 3.725481011227192e-05,
+      "loss": 0.9194,
+      "step": 8437
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6002544419122366,
+      "learning_rate": 3.7251174830074096e-05,
+      "loss": 0.958,
+      "step": 8438
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.414075908154449,
+      "learning_rate": 3.7247539206931056e-05,
+      "loss": 1.457,
+      "step": 8439
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4396641245255104,
+      "learning_rate": 3.724390324294399e-05,
+      "loss": 1.3574,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5244001790297625,
+      "learning_rate": 3.7240266938214076e-05,
+      "loss": 0.9785,
+      "step": 8441
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.224271199045575,
+      "learning_rate": 3.723663029284252e-05,
+      "loss": 1.2861,
+      "step": 8442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.20366184015947,
+      "learning_rate": 3.723299330693052e-05,
+      "loss": 1.2969,
+      "step": 8443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8175404628498493,
+      "learning_rate": 3.72293559805793e-05,
+      "loss": 1.2656,
+      "step": 8444
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.667689296223018,
+      "learning_rate": 3.722571831389009e-05,
+      "loss": 0.7002,
+      "step": 8445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.742228953627142,
+      "learning_rate": 3.722208030696411e-05,
+      "loss": 2.0566,
+      "step": 8446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.371661629149753,
+      "learning_rate": 3.72184419599026e-05,
+      "loss": 0.6416,
+      "step": 8447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6839708153462283,
+      "learning_rate": 3.721480327280684e-05,
+      "loss": 1.207,
+      "step": 8448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.588801981692505,
+      "learning_rate": 3.721116424577808e-05,
+      "loss": 1.7949,
+      "step": 8449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.372077679763399,
+      "learning_rate": 3.7207524878917576e-05,
+      "loss": 1.2832,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7124936257564345,
+      "learning_rate": 3.720388517232663e-05,
+      "loss": 0.9111,
+      "step": 8451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.609293067333255,
+      "learning_rate": 3.720024512610652e-05,
+      "loss": 0.957,
+      "step": 8452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.087045359632605,
+      "learning_rate": 3.719660474035857e-05,
+      "loss": 1.0381,
+      "step": 8453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.05456354616952,
+      "learning_rate": 3.7192964015184064e-05,
+      "loss": 1.1172,
+      "step": 8454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5855622264165996,
+      "learning_rate": 3.718932295068435e-05,
+      "loss": 1.0605,
+      "step": 8455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.72631004060993,
+      "learning_rate": 3.7185681546960724e-05,
+      "loss": 0.8789,
+      "step": 8456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.262643072238598,
+      "learning_rate": 3.7182039804114544e-05,
+      "loss": 1.6016,
+      "step": 8457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.189533480071456,
+      "learning_rate": 3.717839772224715e-05,
+      "loss": 0.9756,
+      "step": 8458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.672719617553313,
+      "learning_rate": 3.717475530145992e-05,
+      "loss": 1.1641,
+      "step": 8459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.162949289812546,
+      "learning_rate": 3.717111254185419e-05,
+      "loss": 1.0811,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5953019757348668,
+      "learning_rate": 3.716746944353137e-05,
+      "loss": 0.4756,
+      "step": 8461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.127452332873736,
+      "learning_rate": 3.716382600659281e-05,
+      "loss": 1.0703,
+      "step": 8462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.231184204130932,
+      "learning_rate": 3.716018223113994e-05,
+      "loss": 1.0151,
+      "step": 8463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8873548995070097,
+      "learning_rate": 3.7156538117274144e-05,
+      "loss": 1.4746,
+      "step": 8464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6921623064966727,
+      "learning_rate": 3.715289366509684e-05,
+      "loss": 0.8608,
+      "step": 8465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2057592997168465,
+      "learning_rate": 3.7149248874709455e-05,
+      "loss": 0.8267,
+      "step": 8466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.919955602926273,
+      "learning_rate": 3.7145603746213415e-05,
+      "loss": 0.9404,
+      "step": 8467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0902858567340603,
+      "learning_rate": 3.7141958279710164e-05,
+      "loss": 1.1602,
+      "step": 8468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2867902603150188,
+      "learning_rate": 3.713831247530117e-05,
+      "loss": 1.0215,
+      "step": 8469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.47593261498915,
+      "learning_rate": 3.713466633308787e-05,
+      "loss": 0.5459,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.708017200039144,
+      "learning_rate": 3.713101985317175e-05,
+      "loss": 1.3564,
+      "step": 8471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.859702145275562,
+      "learning_rate": 3.712737303565429e-05,
+      "loss": 1.6348,
+      "step": 8472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.610140012466058,
+      "learning_rate": 3.712372588063697e-05,
+      "loss": 1.1113,
+      "step": 8473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2751502606814094,
+      "learning_rate": 3.71200783882213e-05,
+      "loss": 1.2891,
+      "step": 8474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.49992370552552,
+      "learning_rate": 3.7116430558508776e-05,
+      "loss": 1.0386,
+      "step": 8475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.343598462629719,
+      "learning_rate": 3.711278239160092e-05,
+      "loss": 1.114,
+      "step": 8476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.135786951727097,
+      "learning_rate": 3.710913388759927e-05,
+      "loss": 0.8789,
+      "step": 8477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9656034483584757,
+      "learning_rate": 3.710548504660535e-05,
+      "loss": 1.3496,
+      "step": 8478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.247995649428587,
+      "learning_rate": 3.710183586872071e-05,
+      "loss": 1.5352,
+      "step": 8479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4323617159111506,
+      "learning_rate": 3.70981863540469e-05,
+      "loss": 1.3496,
+      "step": 8480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.3892700261533495,
+      "learning_rate": 3.709453650268549e-05,
+      "loss": 0.8203,
+      "step": 8481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.84492945880598,
+      "learning_rate": 3.709088631473805e-05,
+      "loss": 0.9355,
+      "step": 8482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3812940418245603,
+      "learning_rate": 3.708723579030618e-05,
+      "loss": 0.7881,
+      "step": 8483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8063770188979396,
+      "learning_rate": 3.708358492949145e-05,
+      "loss": 0.8057,
+      "step": 8484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3201936119781488,
+      "learning_rate": 3.707993373239547e-05,
+      "loss": 1.4364,
+      "step": 8485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8408057417153771,
+      "learning_rate": 3.707628219911985e-05,
+      "loss": 1.2959,
+      "step": 8486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.777546124012156,
+      "learning_rate": 3.707263032976622e-05,
+      "loss": 1.6035,
+      "step": 8487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.33238230352098,
+      "learning_rate": 3.7068978124436194e-05,
+      "loss": 1.1777,
+      "step": 8488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4958953539785025,
+      "learning_rate": 3.706532558323142e-05,
+      "loss": 0.9434,
+      "step": 8489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5319994535159323,
+      "learning_rate": 3.7061672706253545e-05,
+      "loss": 1.4121,
+      "step": 8490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6775481022236893,
+      "learning_rate": 3.7058019493604244e-05,
+      "loss": 1.0566,
+      "step": 8491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.466721775154881,
+      "learning_rate": 3.7054365945385154e-05,
+      "loss": 1.291,
+      "step": 8492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.440605841227312,
+      "learning_rate": 3.705071206169798e-05,
+      "loss": 1.1621,
+      "step": 8493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.131303306154885,
+      "learning_rate": 3.704705784264438e-05,
+      "loss": 1.0107,
+      "step": 8494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.151499831674953,
+      "learning_rate": 3.704340328832607e-05,
+      "loss": 2.0137,
+      "step": 8495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1383759873086268,
+      "learning_rate": 3.703974839884475e-05,
+      "loss": 1.0283,
+      "step": 8496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6886944445231262,
+      "learning_rate": 3.7036093174302145e-05,
+      "loss": 1.1484,
+      "step": 8497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.353026887791121,
+      "learning_rate": 3.7032437614799955e-05,
+      "loss": 1.0522,
+      "step": 8498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.494306429260294,
+      "learning_rate": 3.7028781720439924e-05,
+      "loss": 1.084,
+      "step": 8499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.957610937217212,
+      "learning_rate": 3.70251254913238e-05,
+      "loss": 1.0898,
+      "step": 8500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0143131452313967,
+      "learning_rate": 3.7021468927553326e-05,
+      "loss": 1.2539,
+      "step": 8501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.834625137434689,
+      "learning_rate": 3.701781202923026e-05,
+      "loss": 1.1035,
+      "step": 8502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.458491972206699,
+      "learning_rate": 3.701415479645639e-05,
+      "loss": 0.8828,
+      "step": 8503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4746416506811424,
+      "learning_rate": 3.701049722933348e-05,
+      "loss": 0.9082,
+      "step": 8504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.932746775098489,
+      "learning_rate": 3.7006839327963317e-05,
+      "loss": 1.0723,
+      "step": 8505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.111209960532729,
+      "learning_rate": 3.70031810924477e-05,
+      "loss": 0.6152,
+      "step": 8506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.991420442158275,
+      "learning_rate": 3.699952252288845e-05,
+      "loss": 0.6917,
+      "step": 8507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.638781301707085,
+      "learning_rate": 3.699586361938737e-05,
+      "loss": 0.9414,
+      "step": 8508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9812098814936792,
+      "learning_rate": 3.699220438204628e-05,
+      "loss": 0.999,
+      "step": 8509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7634483356267987,
+      "learning_rate": 3.698854481096704e-05,
+      "loss": 0.791,
+      "step": 8510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0323625862706884,
+      "learning_rate": 3.698488490625146e-05,
+      "loss": 0.7266,
+      "step": 8511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3058426370583534,
+      "learning_rate": 3.6981224668001424e-05,
+      "loss": 0.7759,
+      "step": 8512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3835699159624006,
+      "learning_rate": 3.697756409631878e-05,
+      "loss": 1.2695,
+      "step": 8513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7814217846570912,
+      "learning_rate": 3.69739031913054e-05,
+      "loss": 1.2207,
+      "step": 8514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.137072931533277,
+      "learning_rate": 3.6970241953063176e-05,
+      "loss": 1.1309,
+      "step": 8515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5121300173265166,
+      "learning_rate": 3.696658038169399e-05,
+      "loss": 1.1641,
+      "step": 8516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9439551668692463,
+      "learning_rate": 3.696291847729973e-05,
+      "loss": 1.5371,
+      "step": 8517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.34155180526077,
+      "learning_rate": 3.695925623998234e-05,
+      "loss": 1.2207,
+      "step": 8518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.492818626921356,
+      "learning_rate": 3.69555936698437e-05,
+      "loss": 1.0693,
+      "step": 8519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1868539791307913,
+      "learning_rate": 3.695193076698577e-05,
+      "loss": 0.9194,
+      "step": 8520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3799651596981315,
+      "learning_rate": 3.694826753151047e-05,
+      "loss": 1.5039,
+      "step": 8521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.11458166185263,
+      "learning_rate": 3.6944603963519744e-05,
+      "loss": 1.6348,
+      "step": 8522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.504328844836988,
+      "learning_rate": 3.694094006311555e-05,
+      "loss": 0.9446,
+      "step": 8523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.140442889187034,
+      "learning_rate": 3.693727583039985e-05,
+      "loss": 0.6821,
+      "step": 8524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.780910342617899,
+      "learning_rate": 3.6933611265474646e-05,
+      "loss": 1.6172,
+      "step": 8525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2622891412755215,
+      "learning_rate": 3.6929946368441876e-05,
+      "loss": 1.5898,
+      "step": 8526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.552040875194945,
+      "learning_rate": 3.6926281139403565e-05,
+      "loss": 2.3164,
+      "step": 8527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.247313697230488,
+      "learning_rate": 3.69226155784617e-05,
+      "loss": 1.3887,
+      "step": 8528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.048749357158774,
+      "learning_rate": 3.6918949685718305e-05,
+      "loss": 1.1953,
+      "step": 8529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1192714668348605,
+      "learning_rate": 3.691528346127538e-05,
+      "loss": 0.9596,
+      "step": 8530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0150230750946903,
+      "learning_rate": 3.691161690523498e-05,
+      "loss": 1.3203,
+      "step": 8531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6190901627705006,
+      "learning_rate": 3.690795001769912e-05,
+      "loss": 1.2041,
+      "step": 8532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.143714069193415,
+      "learning_rate": 3.6904282798769866e-05,
+      "loss": 0.9087,
+      "step": 8533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.573804147558626,
+      "learning_rate": 3.690061524854927e-05,
+      "loss": 1.3145,
+      "step": 8534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.573592537749371,
+      "learning_rate": 3.6896947367139395e-05,
+      "loss": 1.5293,
+      "step": 8535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8712256161256797,
+      "learning_rate": 3.689327915464231e-05,
+      "loss": 1.3809,
+      "step": 8536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.425792808463519,
+      "learning_rate": 3.688961061116012e-05,
+      "loss": 1.1865,
+      "step": 8537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.775938821421054,
+      "learning_rate": 3.688594173679489e-05,
+      "loss": 1.0563,
+      "step": 8538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.999845184871993,
+      "learning_rate": 3.6882272531648755e-05,
+      "loss": 1.4531,
+      "step": 8539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.445371282151991,
+      "learning_rate": 3.687860299582381e-05,
+      "loss": 0.689,
+      "step": 8540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.147368675009001,
+      "learning_rate": 3.6874933129422174e-05,
+      "loss": 0.6006,
+      "step": 8541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0423030286749615,
+      "learning_rate": 3.687126293254599e-05,
+      "loss": 0.7896,
+      "step": 8542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.241565790943776,
+      "learning_rate": 3.686759240529739e-05,
+      "loss": 0.9141,
+      "step": 8543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0835235811942887,
+      "learning_rate": 3.6863921547778526e-05,
+      "loss": 1.418,
+      "step": 8544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1297815046043422,
+      "learning_rate": 3.686025036009154e-05,
+      "loss": 1.0293,
+      "step": 8545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.23238139812523,
+      "learning_rate": 3.685657884233863e-05,
+      "loss": 1.3574,
+      "step": 8546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.638455587062821,
+      "learning_rate": 3.685290699462196e-05,
+      "loss": 0.9805,
+      "step": 8547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3789485862593387,
+      "learning_rate": 3.684923481704371e-05,
+      "loss": 1.498,
+      "step": 8548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.158073888775289,
+      "learning_rate": 3.684556230970607e-05,
+      "loss": 1.0137,
+      "step": 8549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7345241448136866,
+      "learning_rate": 3.684188947271127e-05,
+      "loss": 1.1152,
+      "step": 8550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.145584709307002,
+      "learning_rate": 3.683821630616149e-05,
+      "loss": 0.8604,
+      "step": 8551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.8437590568211695,
+      "learning_rate": 3.6834542810158977e-05,
+      "loss": 1.1167,
+      "step": 8552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6437705000855884,
+      "learning_rate": 3.683086898480595e-05,
+      "loss": 0.9082,
+      "step": 8553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.2278060460089275,
+      "learning_rate": 3.682719483020466e-05,
+      "loss": 1.1279,
+      "step": 8554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6205131695042883,
+      "learning_rate": 3.6823520346457365e-05,
+      "loss": 0.7869,
+      "step": 8555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.260874219956598,
+      "learning_rate": 3.68198455336663e-05,
+      "loss": 0.9248,
+      "step": 8556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.849477512584818,
+      "learning_rate": 3.681617039193374e-05,
+      "loss": 0.3154,
+      "step": 8557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6035498740597056,
+      "learning_rate": 3.6812494921361976e-05,
+      "loss": 1.041,
+      "step": 8558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5698017447850394,
+      "learning_rate": 3.6808819122053285e-05,
+      "loss": 1.1108,
+      "step": 8559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7602643144898185,
+      "learning_rate": 3.6805142994109965e-05,
+      "loss": 0.8916,
+      "step": 8560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.304114112001723,
+      "learning_rate": 3.6801466537634326e-05,
+      "loss": 1.7109,
+      "step": 8561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6945735523224474,
+      "learning_rate": 3.6797789752728675e-05,
+      "loss": 1.9336,
+      "step": 8562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2058807885407528,
+      "learning_rate": 3.679411263949534e-05,
+      "loss": 1.4434,
+      "step": 8563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.567376196220008,
+      "learning_rate": 3.6790435198036644e-05,
+      "loss": 1.2812,
+      "step": 8564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.154672647288866,
+      "learning_rate": 3.678675742845494e-05,
+      "loss": 0.9375,
+      "step": 8565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2518632389510644,
+      "learning_rate": 3.678307933085258e-05,
+      "loss": 0.9531,
+      "step": 8566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9538589218433593,
+      "learning_rate": 3.6779400905331917e-05,
+      "loss": 0.8003,
+      "step": 8567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.378123595588319,
+      "learning_rate": 3.677572215199531e-05,
+      "loss": 1.4141,
+      "step": 8568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2451157007797384,
+      "learning_rate": 3.677204307094516e-05,
+      "loss": 1.0166,
+      "step": 8569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.387195452411466,
+      "learning_rate": 3.6768363662283844e-05,
+      "loss": 1.0176,
+      "step": 8570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6441255199053333,
+      "learning_rate": 3.676468392611376e-05,
+      "loss": 0.9062,
+      "step": 8571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.103028941718792,
+      "learning_rate": 3.6761003862537314e-05,
+      "loss": 0.9746,
+      "step": 8572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1792011333630676,
+      "learning_rate": 3.675732347165691e-05,
+      "loss": 1.2148,
+      "step": 8573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.986689953454262,
+      "learning_rate": 3.675364275357499e-05,
+      "loss": 0.6348,
+      "step": 8574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9041913031942843,
+      "learning_rate": 3.674996170839397e-05,
+      "loss": 1.1377,
+      "step": 8575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.824592027490897,
+      "learning_rate": 3.67462803362163e-05,
+      "loss": 0.6758,
+      "step": 8576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.561629957042793,
+      "learning_rate": 3.6742598637144435e-05,
+      "loss": 0.9326,
+      "step": 8577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8549952916843053,
+      "learning_rate": 3.6738916611280814e-05,
+      "loss": 0.9912,
+      "step": 8578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.239335997676573,
+      "learning_rate": 3.6735234258727935e-05,
+      "loss": 0.5911,
+      "step": 8579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.532120858894953,
+      "learning_rate": 3.673155157958827e-05,
+      "loss": 0.9951,
+      "step": 8580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1648331068731492,
+      "learning_rate": 3.6727868573964294e-05,
+      "loss": 1.3047,
+      "step": 8581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.234094138877865,
+      "learning_rate": 3.672418524195851e-05,
+      "loss": 1.0889,
+      "step": 8582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9005582312119995,
+      "learning_rate": 3.6720501583673414e-05,
+      "loss": 1.0723,
+      "step": 8583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3126597815790277,
+      "learning_rate": 3.671681759921155e-05,
+      "loss": 0.7139,
+      "step": 8584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.209496981011316,
+      "learning_rate": 3.6713133288675404e-05,
+      "loss": 0.8467,
+      "step": 8585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.843306353017481,
+      "learning_rate": 3.6709448652167545e-05,
+      "loss": 1.1289,
+      "step": 8586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.969926814529421,
+      "learning_rate": 3.670576368979048e-05,
+      "loss": 0.8037,
+      "step": 8587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.403730213165236,
+      "learning_rate": 3.670207840164678e-05,
+      "loss": 0.9961,
+      "step": 8588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.585724236951676,
+      "learning_rate": 3.669839278783901e-05,
+      "loss": 0.9902,
+      "step": 8589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.376454792651699,
+      "learning_rate": 3.669470684846973e-05,
+      "loss": 1.2031,
+      "step": 8590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2607932297153703,
+      "learning_rate": 3.669102058364151e-05,
+      "loss": 0.3445,
+      "step": 8591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.460760314401353,
+      "learning_rate": 3.668733399345695e-05,
+      "loss": 1.2969,
+      "step": 8592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.811831341154103,
+      "learning_rate": 3.668364707801866e-05,
+      "loss": 0.9365,
+      "step": 8593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4484751179330386,
+      "learning_rate": 3.667995983742921e-05,
+      "loss": 1.2031,
+      "step": 8594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3121451028215043,
+      "learning_rate": 3.667627227179123e-05,
+      "loss": 1.8301,
+      "step": 8595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8292150915364591,
+      "learning_rate": 3.6672584381207356e-05,
+      "loss": 0.5411,
+      "step": 8596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.903837447143584,
+      "learning_rate": 3.666889616578021e-05,
+      "loss": 1.3457,
+      "step": 8597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.429863232257547,
+      "learning_rate": 3.666520762561243e-05,
+      "loss": 0.7432,
+      "step": 8598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.984953659439125,
+      "learning_rate": 3.666151876080667e-05,
+      "loss": 1.5488,
+      "step": 8599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.67390947661128,
+      "learning_rate": 3.66578295714656e-05,
+      "loss": 1.2197,
+      "step": 8600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0642285820877806,
+      "learning_rate": 3.665414005769186e-05,
+      "loss": 1.0771,
+      "step": 8601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.296077882096102,
+      "learning_rate": 3.665045021958816e-05,
+      "loss": 0.8687,
+      "step": 8602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0274057250206323,
+      "learning_rate": 3.664676005725717e-05,
+      "loss": 1.0864,
+      "step": 8603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.14248860043785,
+      "learning_rate": 3.664306957080159e-05,
+      "loss": 1.498,
+      "step": 8604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3025549101614264,
+      "learning_rate": 3.663937876032413e-05,
+      "loss": 0.6099,
+      "step": 8605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8132599870479327,
+      "learning_rate": 3.663568762592748e-05,
+      "loss": 1.4648,
+      "step": 8606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0241832355610554,
+      "learning_rate": 3.66319961677144e-05,
+      "loss": 1.0317,
+      "step": 8607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.214386100515116,
+      "learning_rate": 3.6628304385787594e-05,
+      "loss": 0.7832,
+      "step": 8608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3801041016130746,
+      "learning_rate": 3.662461228024981e-05,
+      "loss": 1.0742,
+      "step": 8609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.992499431662599,
+      "learning_rate": 3.66209198512038e-05,
+      "loss": 1.3203,
+      "step": 8610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.215768070405951,
+      "learning_rate": 3.661722709875232e-05,
+      "loss": 1.1797,
+      "step": 8611
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0421690679534135,
+      "learning_rate": 3.661353402299814e-05,
+      "loss": 1.1641,
+      "step": 8612
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.303938300995428,
+      "learning_rate": 3.6609840624044035e-05,
+      "loss": 1.6523,
+      "step": 8613
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.154736416090643,
+      "learning_rate": 3.6606146901992796e-05,
+      "loss": 0.9712,
+      "step": 8614
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.50174756958951,
+      "learning_rate": 3.6602452856947206e-05,
+      "loss": 1.0459,
+      "step": 8615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4139066479196343,
+      "learning_rate": 3.659875848901008e-05,
+      "loss": 0.4878,
+      "step": 8616
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4276661207282375,
+      "learning_rate": 3.6595063798284226e-05,
+      "loss": 0.6372,
+      "step": 8617
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.831502809747847,
+      "learning_rate": 3.659136878487247e-05,
+      "loss": 1.3593,
+      "step": 8618
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.790093539248719,
+      "learning_rate": 3.658767344887764e-05,
+      "loss": 0.8218,
+      "step": 8619
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.090814595747075,
+      "learning_rate": 3.658397779040257e-05,
+      "loss": 0.9648,
+      "step": 8620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.473106062129542,
+      "learning_rate": 3.658028180955012e-05,
+      "loss": 1.519,
+      "step": 8621
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9919669366325463,
+      "learning_rate": 3.657658550642314e-05,
+      "loss": 1.4492,
+      "step": 8622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.984834965116486,
+      "learning_rate": 3.6572888881124496e-05,
+      "loss": 0.8574,
+      "step": 8623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.58800011834623,
+      "learning_rate": 3.6569191933757064e-05,
+      "loss": 1.1431,
+      "step": 8624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1696077217248773,
+      "learning_rate": 3.6565494664423734e-05,
+      "loss": 1.4004,
+      "step": 8625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.028611993032667,
+      "learning_rate": 3.65617970732274e-05,
+      "loss": 0.7969,
+      "step": 8626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.485352405429955,
+      "learning_rate": 3.655809916027095e-05,
+      "loss": 1.3926,
+      "step": 8627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3445356640322133,
+      "learning_rate": 3.655440092565732e-05,
+      "loss": 0.6001,
+      "step": 8628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.651848950533441,
+      "learning_rate": 3.65507023694894e-05,
+      "loss": 0.9609,
+      "step": 8629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.804915479994816,
+      "learning_rate": 3.654700349187014e-05,
+      "loss": 1.334,
+      "step": 8630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.580767350296879,
+      "learning_rate": 3.654330429290248e-05,
+      "loss": 1.4785,
+      "step": 8631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.624043233629014,
+      "learning_rate": 3.653960477268935e-05,
+      "loss": 0.6982,
+      "step": 8632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1712770697100328,
+      "learning_rate": 3.6535904931333726e-05,
+      "loss": 0.8701,
+      "step": 8633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.947183201912429,
+      "learning_rate": 3.653220476893856e-05,
+      "loss": 0.7148,
+      "step": 8634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.537927555838922,
+      "learning_rate": 3.652850428560684e-05,
+      "loss": 1.3496,
+      "step": 8635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.943607657176166,
+      "learning_rate": 3.652480348144153e-05,
+      "loss": 1.1763,
+      "step": 8636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2892706974898136,
+      "learning_rate": 3.652110235654563e-05,
+      "loss": 0.8633,
+      "step": 8637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.576784374176138,
+      "learning_rate": 3.6517400911022136e-05,
+      "loss": 0.6937,
+      "step": 8638
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.670591603250879,
+      "learning_rate": 3.651369914497407e-05,
+      "loss": 1.2891,
+      "step": 8639
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.410063176094706,
+      "learning_rate": 3.650999705850445e-05,
+      "loss": 1.0498,
+      "step": 8640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.455834350890217,
+      "learning_rate": 3.6506294651716294e-05,
+      "loss": 1.2676,
+      "step": 8641
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.180616238462614,
+      "learning_rate": 3.650259192471264e-05,
+      "loss": 1.2178,
+      "step": 8642
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.414299616850215,
+      "learning_rate": 3.649888887759653e-05,
+      "loss": 1.1641,
+      "step": 8643
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1739900800024725,
+      "learning_rate": 3.649518551047103e-05,
+      "loss": 1.0723,
+      "step": 8644
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.795702969224359,
+      "learning_rate": 3.64914818234392e-05,
+      "loss": 1.124,
+      "step": 8645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.489769007596844,
+      "learning_rate": 3.64877778166041e-05,
+      "loss": 0.7432,
+      "step": 8646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.286220588750048,
+      "learning_rate": 3.6484073490068826e-05,
+      "loss": 0.8251,
+      "step": 8647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.777467222100131,
+      "learning_rate": 3.648036884393646e-05,
+      "loss": 1.4512,
+      "step": 8648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.872218585668463,
+      "learning_rate": 3.647666387831011e-05,
+      "loss": 1.2598,
+      "step": 8649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.07961084782648,
+      "learning_rate": 3.647295859329287e-05,
+      "loss": 1.1948,
+      "step": 8650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.741742715022816,
+      "learning_rate": 3.6469252988987865e-05,
+      "loss": 1.4551,
+      "step": 8651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.3682287950365115,
+      "learning_rate": 3.646554706549821e-05,
+      "loss": 1.3467,
+      "step": 8652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.267718360568515,
+      "learning_rate": 3.646184082292706e-05,
+      "loss": 1.2695,
+      "step": 8653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7334423981593567,
+      "learning_rate": 3.6458134261377544e-05,
+      "loss": 1.2275,
+      "step": 8654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.0134856513063335,
+      "learning_rate": 3.6454427380952805e-05,
+      "loss": 1.4023,
+      "step": 8655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.84099198864368,
+      "learning_rate": 3.6450720181756026e-05,
+      "loss": 1.5952,
+      "step": 8656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4010268160566928,
+      "learning_rate": 3.6447012663890355e-05,
+      "loss": 1.6426,
+      "step": 8657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.507351909623057,
+      "learning_rate": 3.644330482745899e-05,
+      "loss": 1.0713,
+      "step": 8658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0049396220661713,
+      "learning_rate": 3.643959667256511e-05,
+      "loss": 0.9072,
+      "step": 8659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.103024293080578,
+      "learning_rate": 3.6435888199311916e-05,
+      "loss": 1.2435,
+      "step": 8660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9726288201448825,
+      "learning_rate": 3.6432179407802606e-05,
+      "loss": 1.8262,
+      "step": 8661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.238267555714033,
+      "learning_rate": 3.6428470298140395e-05,
+      "loss": 1.126,
+      "step": 8662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.373802686350655,
+      "learning_rate": 3.642476087042851e-05,
+      "loss": 1.2383,
+      "step": 8663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.462408966072589,
+      "learning_rate": 3.642105112477018e-05,
+      "loss": 1.1094,
+      "step": 8664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.180245531957178,
+      "learning_rate": 3.6417341061268644e-05,
+      "loss": 1.5703,
+      "step": 8665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.073781707818192,
+      "learning_rate": 3.6413630680027156e-05,
+      "loss": 1.0874,
+      "step": 8666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3932151973899183,
+      "learning_rate": 3.640991998114898e-05,
+      "loss": 1.2285,
+      "step": 8667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.729131872487202,
+      "learning_rate": 3.640620896473736e-05,
+      "loss": 1.2715,
+      "step": 8668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0373779963100183,
+      "learning_rate": 3.6402497630895605e-05,
+      "loss": 1.8086,
+      "step": 8669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9876238573503846,
+      "learning_rate": 3.639878597972697e-05,
+      "loss": 0.9941,
+      "step": 8670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.543105218481861,
+      "learning_rate": 3.639507401133477e-05,
+      "loss": 1.3652,
+      "step": 8671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.772364493452344,
+      "learning_rate": 3.6391361725822294e-05,
+      "loss": 1.0693,
+      "step": 8672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8889425794388717,
+      "learning_rate": 3.638764912329286e-05,
+      "loss": 1.4062,
+      "step": 8673
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0713603243421894,
+      "learning_rate": 3.638393620384979e-05,
+      "loss": 0.5889,
+      "step": 8674
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8007941586906837,
+      "learning_rate": 3.63802229675964e-05,
+      "loss": 1.1357,
+      "step": 8675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.902909149807703,
+      "learning_rate": 3.637650941463604e-05,
+      "loss": 0.8395,
+      "step": 8676
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.61904513350695,
+      "learning_rate": 3.637279554507207e-05,
+      "loss": 1.4268,
+      "step": 8677
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6428144959650512,
+      "learning_rate": 3.636908135900782e-05,
+      "loss": 0.9287,
+      "step": 8678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.171643032063446,
+      "learning_rate": 3.636536685654666e-05,
+      "loss": 1.1992,
+      "step": 8679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0387761390531565,
+      "learning_rate": 3.636165203779197e-05,
+      "loss": 1.3691,
+      "step": 8680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8957416510762544,
+      "learning_rate": 3.635793690284713e-05,
+      "loss": 1.1387,
+      "step": 8681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1852299491993254,
+      "learning_rate": 3.6354221451815535e-05,
+      "loss": 0.9141,
+      "step": 8682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4010371430879185,
+      "learning_rate": 3.635050568480057e-05,
+      "loss": 0.877,
+      "step": 8683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.790612323144512,
+      "learning_rate": 3.634678960190565e-05,
+      "loss": 0.5717,
+      "step": 8684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.600849304722962,
+      "learning_rate": 3.63430732032342e-05,
+      "loss": 2.4414,
+      "step": 8685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4774725652844394,
+      "learning_rate": 3.633935648888965e-05,
+      "loss": 1.0898,
+      "step": 8686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1551853813570507,
+      "learning_rate": 3.633563945897541e-05,
+      "loss": 1.0537,
+      "step": 8687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0658343556739993,
+      "learning_rate": 3.633192211359495e-05,
+      "loss": 1.8145,
+      "step": 8688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7340718128509223,
+      "learning_rate": 3.63282044528517e-05,
+      "loss": 0.8701,
+      "step": 8689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.621801082985756,
+      "learning_rate": 3.632448647684914e-05,
+      "loss": 0.8242,
+      "step": 8690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2931257144900417,
+      "learning_rate": 3.632076818569073e-05,
+      "loss": 1.0312,
+      "step": 8691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.595224482881923,
+      "learning_rate": 3.631704957947996e-05,
+      "loss": 2.0615,
+      "step": 8692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8137845814502014,
+      "learning_rate": 3.631333065832028e-05,
+      "loss": 1.1904,
+      "step": 8693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.351060665236051,
+      "learning_rate": 3.630961142231524e-05,
+      "loss": 1.2266,
+      "step": 8694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.064463481899842,
+      "learning_rate": 3.63058918715683e-05,
+      "loss": 1.3613,
+      "step": 8695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.24670872812892,
+      "learning_rate": 3.630217200618301e-05,
+      "loss": 1.7617,
+      "step": 8696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.964004489038935,
+      "learning_rate": 3.6298451826262857e-05,
+      "loss": 0.5682,
+      "step": 8697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.535003610491768,
+      "learning_rate": 3.62947313319114e-05,
+      "loss": 0.791,
+      "step": 8698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.169614974464514,
+      "learning_rate": 3.629101052323216e-05,
+      "loss": 0.8057,
+      "step": 8699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.464707416758452,
+      "learning_rate": 3.628728940032871e-05,
+      "loss": 0.9834,
+      "step": 8700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.775616115185631,
+      "learning_rate": 3.628356796330458e-05,
+      "loss": 1.2227,
+      "step": 8701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7294482505687774,
+      "learning_rate": 3.627984621226334e-05,
+      "loss": 1.2266,
+      "step": 8702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.38854922811451,
+      "learning_rate": 3.627612414730858e-05,
+      "loss": 0.8652,
+      "step": 8703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.510428801912178,
+      "learning_rate": 3.627240176854387e-05,
+      "loss": 0.9648,
+      "step": 8704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.496005491030595,
+      "learning_rate": 3.626867907607282e-05,
+      "loss": 0.688,
+      "step": 8705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.684021740114496,
+      "learning_rate": 3.6264956069999005e-05,
+      "loss": 0.8418,
+      "step": 8706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.2727263203093715,
+      "learning_rate": 3.626123275042605e-05,
+      "loss": 1.0898,
+      "step": 8707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1522421700380256,
+      "learning_rate": 3.625750911745758e-05,
+      "loss": 0.8623,
+      "step": 8708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1156505342621386,
+      "learning_rate": 3.625378517119721e-05,
+      "loss": 1.293,
+      "step": 8709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4389546284351886,
+      "learning_rate": 3.6250060911748586e-05,
+      "loss": 1.2852,
+      "step": 8710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.225871764765373,
+      "learning_rate": 3.624633633921534e-05,
+      "loss": 1.5957,
+      "step": 8711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5369952897872774,
+      "learning_rate": 3.624261145370113e-05,
+      "loss": 1.3672,
+      "step": 8712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.019112955193274,
+      "learning_rate": 3.6238886255309626e-05,
+      "loss": 1.5645,
+      "step": 8713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.1770438222245,
+      "learning_rate": 3.6235160744144484e-05,
+      "loss": 1.603,
+      "step": 8714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.953555387399655,
+      "learning_rate": 3.62314349203094e-05,
+      "loss": 0.9033,
+      "step": 8715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.797303495383269,
+      "learning_rate": 3.6227708783908055e-05,
+      "loss": 1.1289,
+      "step": 8716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7902838871904576,
+      "learning_rate": 3.622398233504414e-05,
+      "loss": 1.251,
+      "step": 8717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4396329331282787,
+      "learning_rate": 3.622025557382138e-05,
+      "loss": 0.3374,
+      "step": 8718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.604965138412599,
+      "learning_rate": 3.621652850034346e-05,
+      "loss": 1.0464,
+      "step": 8719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.327153486864207,
+      "learning_rate": 3.621280111471412e-05,
+      "loss": 1.6836,
+      "step": 8720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.882562393876296,
+      "learning_rate": 3.620907341703709e-05,
+      "loss": 1.9062,
+      "step": 8721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2428355844644132,
+      "learning_rate": 3.620534540741612e-05,
+      "loss": 1.0869,
+      "step": 8722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8605420593289097,
+      "learning_rate": 3.6201617085954934e-05,
+      "loss": 1.0645,
+      "step": 8723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.326355126991553,
+      "learning_rate": 3.619788845275732e-05,
+      "loss": 1.1992,
+      "step": 8724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.437852188106866,
+      "learning_rate": 3.6194159507927014e-05,
+      "loss": 2.2246,
+      "step": 8725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.393618429796026,
+      "learning_rate": 3.619043025156782e-05,
+      "loss": 0.7637,
+      "step": 8726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7348536739258216,
+      "learning_rate": 3.618670068378349e-05,
+      "loss": 0.9277,
+      "step": 8727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7945945532614611,
+      "learning_rate": 3.618297080467785e-05,
+      "loss": 0.6377,
+      "step": 8728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.845105529329705,
+      "learning_rate": 3.617924061435468e-05,
+      "loss": 1.001,
+      "step": 8729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.966715846148913,
+      "learning_rate": 3.61755101129178e-05,
+      "loss": 0.7495,
+      "step": 8730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.281221709888853,
+      "learning_rate": 3.6171779300471014e-05,
+      "loss": 1.0125,
+      "step": 8731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4349250886896945,
+      "learning_rate": 3.616804817711816e-05,
+      "loss": 0.8184,
+      "step": 8732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7733107846822826,
+      "learning_rate": 3.616431674296308e-05,
+      "loss": 0.6611,
+      "step": 8733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6040966990926235,
+      "learning_rate": 3.6160584998109605e-05,
+      "loss": 1.2441,
+      "step": 8734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.237498577646562,
+      "learning_rate": 3.615685294266159e-05,
+      "loss": 1.2129,
+      "step": 8735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3528990314775666,
+      "learning_rate": 3.61531205767229e-05,
+      "loss": 1.8613,
+      "step": 8736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1595380840189238,
+      "learning_rate": 3.6149387900397406e-05,
+      "loss": 0.7493,
+      "step": 8737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.161420325975953,
+      "learning_rate": 3.6145654913788985e-05,
+      "loss": 1.0332,
+      "step": 8738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.838771948383096,
+      "learning_rate": 3.614192161700153e-05,
+      "loss": 1.5508,
+      "step": 8739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.843471592931376,
+      "learning_rate": 3.6138188010138916e-05,
+      "loss": 0.9365,
+      "step": 8740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.514661142261595,
+      "learning_rate": 3.613445409330508e-05,
+      "loss": 1.1211,
+      "step": 8741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.541741469069417,
+      "learning_rate": 3.6130719866603916e-05,
+      "loss": 0.5811,
+      "step": 8742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.174997825393567,
+      "learning_rate": 3.612698533013936e-05,
+      "loss": 0.6216,
+      "step": 8743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.333947543698361,
+      "learning_rate": 3.612325048401531e-05,
+      "loss": 0.6562,
+      "step": 8744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.388800030418442,
+      "learning_rate": 3.611951532833574e-05,
+      "loss": 1.0879,
+      "step": 8745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6098554547396726,
+      "learning_rate": 3.611577986320459e-05,
+      "loss": 1.6855,
+      "step": 8746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2687471121777527,
+      "learning_rate": 3.61120440887258e-05,
+      "loss": 1.2686,
+      "step": 8747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.27445250732352,
+      "learning_rate": 3.610830800500335e-05,
+      "loss": 1.0977,
+      "step": 8748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.300006037884465,
+      "learning_rate": 3.610457161214121e-05,
+      "loss": 1.373,
+      "step": 8749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2336640244575947,
+      "learning_rate": 3.6100834910243375e-05,
+      "loss": 1.1934,
+      "step": 8750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6661781121786663,
+      "learning_rate": 3.609709789941381e-05,
+      "loss": 1.2012,
+      "step": 8751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.218026164011485,
+      "learning_rate": 3.609336057975654e-05,
+      "loss": 1.1543,
+      "step": 8752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.239155947125594,
+      "learning_rate": 3.608962295137555e-05,
+      "loss": 1.5078,
+      "step": 8753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.508634760023481,
+      "learning_rate": 3.6085885014374866e-05,
+      "loss": 0.8574,
+      "step": 8754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7238088436694317,
+      "learning_rate": 3.608214676885852e-05,
+      "loss": 0.9746,
+      "step": 8755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.143154720056364,
+      "learning_rate": 3.607840821493054e-05,
+      "loss": 1.1055,
+      "step": 8756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0544881536116977,
+      "learning_rate": 3.607466935269497e-05,
+      "loss": 1.3223,
+      "step": 8757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.633229027464946,
+      "learning_rate": 3.607093018225586e-05,
+      "loss": 0.9033,
+      "step": 8758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6587724730428912,
+      "learning_rate": 3.606719070371727e-05,
+      "loss": 0.6966,
+      "step": 8759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.759416075471778,
+      "learning_rate": 3.6063450917183275e-05,
+      "loss": 0.6387,
+      "step": 8760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.975099425601882,
+      "learning_rate": 3.605971082275793e-05,
+      "loss": 1.0762,
+      "step": 8761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5946642168934715,
+      "learning_rate": 3.6055970420545346e-05,
+      "loss": 0.7627,
+      "step": 8762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2149766885465056,
+      "learning_rate": 3.6052229710649604e-05,
+      "loss": 0.6875,
+      "step": 8763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3115168363076424,
+      "learning_rate": 3.6048488693174806e-05,
+      "loss": 1.0244,
+      "step": 8764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7479192683218123,
+      "learning_rate": 3.6044747368225076e-05,
+      "loss": 0.9082,
+      "step": 8765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4339375015861022,
+      "learning_rate": 3.604100573590452e-05,
+      "loss": 1.2969,
+      "step": 8766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5379384931670455,
+      "learning_rate": 3.6037263796317256e-05,
+      "loss": 0.7559,
+      "step": 8767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1867821576991866,
+      "learning_rate": 3.603352154956744e-05,
+      "loss": 1.1406,
+      "step": 8768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6634331012829597,
+      "learning_rate": 3.602977899575922e-05,
+      "loss": 1.3535,
+      "step": 8769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.676312332777977,
+      "learning_rate": 3.602603613499673e-05,
+      "loss": 1.377,
+      "step": 8770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.102305287315084,
+      "learning_rate": 3.602229296738415e-05,
+      "loss": 1.0605,
+      "step": 8771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4804158661631845,
+      "learning_rate": 3.601854949302563e-05,
+      "loss": 0.9414,
+      "step": 8772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.126056189135495,
+      "learning_rate": 3.601480571202538e-05,
+      "loss": 1.2227,
+      "step": 8773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0179511867390354,
+      "learning_rate": 3.601106162448756e-05,
+      "loss": 0.8198,
+      "step": 8774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.105595583095592,
+      "learning_rate": 3.600731723051639e-05,
+      "loss": 1.416,
+      "step": 8775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.085870932804095,
+      "learning_rate": 3.6003572530216054e-05,
+      "loss": 1.1045,
+      "step": 8776
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.473295021253122,
+      "learning_rate": 3.599982752369078e-05,
+      "loss": 1.2539,
+      "step": 8777
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.267389177722189,
+      "learning_rate": 3.5996082211044774e-05,
+      "loss": 1.4844,
+      "step": 8778
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1631220898502574,
+      "learning_rate": 3.5992336592382286e-05,
+      "loss": 1.2363,
+      "step": 8779
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.241174874607735,
+      "learning_rate": 3.598859066780754e-05,
+      "loss": 1.123,
+      "step": 8780
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9674401542936264,
+      "learning_rate": 3.598484443742479e-05,
+      "loss": 0.4595,
+      "step": 8781
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.550917617400723,
+      "learning_rate": 3.598109790133829e-05,
+      "loss": 1.5527,
+      "step": 8782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6436158346848515,
+      "learning_rate": 3.597735105965231e-05,
+      "loss": 1.5566,
+      "step": 8783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1782128599554147,
+      "learning_rate": 3.597360391247111e-05,
+      "loss": 1.6094,
+      "step": 8784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.663763482548834,
+      "learning_rate": 3.5969856459899e-05,
+      "loss": 0.999,
+      "step": 8785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.475632990823055,
+      "learning_rate": 3.5966108702040236e-05,
+      "loss": 0.9019,
+      "step": 8786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6457296974613793,
+      "learning_rate": 3.596236063899913e-05,
+      "loss": 0.2711,
+      "step": 8787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.739616488945099,
+      "learning_rate": 3.5958612270879996e-05,
+      "loss": 1.7168,
+      "step": 8788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.403721459309059,
+      "learning_rate": 3.595486359778714e-05,
+      "loss": 0.8271,
+      "step": 8789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7489649297824554,
+      "learning_rate": 3.595111461982489e-05,
+      "loss": 1.0776,
+      "step": 8790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9207799629635969,
+      "learning_rate": 3.5947365337097576e-05,
+      "loss": 0.9243,
+      "step": 8791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.001219625493488,
+      "learning_rate": 3.5943615749709546e-05,
+      "loss": 1.3047,
+      "step": 8792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.288687027805677,
+      "learning_rate": 3.593986585776514e-05,
+      "loss": 0.8745,
+      "step": 8793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.781937094791194,
+      "learning_rate": 3.593611566136873e-05,
+      "loss": 1.0039,
+      "step": 8794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.222743898853151,
+      "learning_rate": 3.5932365160624666e-05,
+      "loss": 1.4004,
+      "step": 8795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.5566291259217575,
+      "learning_rate": 3.592861435563733e-05,
+      "loss": 1.9609,
+      "step": 8796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0013661452908647,
+      "learning_rate": 3.592486324651111e-05,
+      "loss": 1.4922,
+      "step": 8797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.029052017622768,
+      "learning_rate": 3.592111183335039e-05,
+      "loss": 1.4951,
+      "step": 8798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.667347945875292,
+      "learning_rate": 3.5917360116259575e-05,
+      "loss": 1.5078,
+      "step": 8799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0218262609064044,
+      "learning_rate": 3.591360809534307e-05,
+      "loss": 1.1846,
+      "step": 8800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.3860151399838045,
+      "learning_rate": 3.5909855770705294e-05,
+      "loss": 0.9004,
+      "step": 8801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.625031788951397,
+      "learning_rate": 3.5906103142450674e-05,
+      "loss": 0.7363,
+      "step": 8802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.1593988225099245,
+      "learning_rate": 3.5902350210683644e-05,
+      "loss": 0.6875,
+      "step": 8803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9847355312325035,
+      "learning_rate": 3.589859697550865e-05,
+      "loss": 0.741,
+      "step": 8804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.371574642176954,
+      "learning_rate": 3.589484343703014e-05,
+      "loss": 1.459,
+      "step": 8805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.6082830614819255,
+      "learning_rate": 3.5891089595352566e-05,
+      "loss": 0.4717,
+      "step": 8806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4339721610041147,
+      "learning_rate": 3.58873354505804e-05,
+      "loss": 0.8857,
+      "step": 8807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2485189731320374,
+      "learning_rate": 3.588358100281813e-05,
+      "loss": 1.1094,
+      "step": 8808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.532937521649713,
+      "learning_rate": 3.5879826252170226e-05,
+      "loss": 1.4629,
+      "step": 8809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.283209805819348,
+      "learning_rate": 3.5876071198741185e-05,
+      "loss": 0.8955,
+      "step": 8810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.488035869957637,
+      "learning_rate": 3.587231584263552e-05,
+      "loss": 1.9336,
+      "step": 8811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.24657376960005,
+      "learning_rate": 3.5868560183957716e-05,
+      "loss": 0.9131,
+      "step": 8812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.447911742726237,
+      "learning_rate": 3.586480422281232e-05,
+      "loss": 0.9346,
+      "step": 8813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.143115078070705,
+      "learning_rate": 3.5861047959303833e-05,
+      "loss": 0.9932,
+      "step": 8814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2507040288288986,
+      "learning_rate": 3.5857291393536815e-05,
+      "loss": 1.248,
+      "step": 8815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.532580555786017,
+      "learning_rate": 3.585353452561579e-05,
+      "loss": 1.2754,
+      "step": 8816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1362828676609333,
+      "learning_rate": 3.584977735564532e-05,
+      "loss": 1.0557,
+      "step": 8817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5190850387642008,
+      "learning_rate": 3.5846019883729966e-05,
+      "loss": 1.1348,
+      "step": 8818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.004665296340361,
+      "learning_rate": 3.584226210997429e-05,
+      "loss": 1.1797,
+      "step": 8819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.706568151395236,
+      "learning_rate": 3.583850403448287e-05,
+      "loss": 1.3262,
+      "step": 8820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.051803827656578,
+      "learning_rate": 3.58347456573603e-05,
+      "loss": 0.9326,
+      "step": 8821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.487860894989849,
+      "learning_rate": 3.583098697871117e-05,
+      "loss": 0.7832,
+      "step": 8822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.676931047870395,
+      "learning_rate": 3.5827227998640076e-05,
+      "loss": 1.3594,
+      "step": 8823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.83987716382294,
+      "learning_rate": 3.582346871725164e-05,
+      "loss": 0.8213,
+      "step": 8824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.577226701930688,
+      "learning_rate": 3.581970913465047e-05,
+      "loss": 1.791,
+      "step": 8825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.862863670437925,
+      "learning_rate": 3.581594925094122e-05,
+      "loss": 0.9863,
+      "step": 8826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8322408664342955,
+      "learning_rate": 3.581218906622849e-05,
+      "loss": 1.0918,
+      "step": 8827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.346543765251544,
+      "learning_rate": 3.580842858061693e-05,
+      "loss": 1.3145,
+      "step": 8828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.432458942393147,
+      "learning_rate": 3.580466779421121e-05,
+      "loss": 0.7441,
+      "step": 8829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3501198248997786,
+      "learning_rate": 3.580090670711599e-05,
+      "loss": 1.4219,
+      "step": 8830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5524442664029436,
+      "learning_rate": 3.579714531943593e-05,
+      "loss": 0.9209,
+      "step": 8831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.281763306441116,
+      "learning_rate": 3.5793383631275714e-05,
+      "loss": 1.1001,
+      "step": 8832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.25975800586431,
+      "learning_rate": 3.578962164274001e-05,
+      "loss": 1.6641,
+      "step": 8833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.200606147986032,
+      "learning_rate": 3.578585935393355e-05,
+      "loss": 1.2637,
+      "step": 8834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8995403649022986,
+      "learning_rate": 3.5782096764961004e-05,
+      "loss": 1.123,
+      "step": 8835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8383393702347046,
+      "learning_rate": 3.57783338759271e-05,
+      "loss": 1.0598,
+      "step": 8836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.330199978296615,
+      "learning_rate": 3.577457068693655e-05,
+      "loss": 1.0879,
+      "step": 8837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.090629261734435,
+      "learning_rate": 3.5770807198094076e-05,
+      "loss": 0.8398,
+      "step": 8838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.284855645125372,
+      "learning_rate": 3.576704340950443e-05,
+      "loss": 1.1172,
+      "step": 8839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.216490165203233,
+      "learning_rate": 3.5763279321272346e-05,
+      "loss": 0.6035,
+      "step": 8840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1736774303167077,
+      "learning_rate": 3.575951493350258e-05,
+      "loss": 0.9873,
+      "step": 8841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9838494442244134,
+      "learning_rate": 3.57557502462999e-05,
+      "loss": 0.8525,
+      "step": 8842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.606553068599032,
+      "learning_rate": 3.5751985259769065e-05,
+      "loss": 1.25,
+      "step": 8843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.25399842423657,
+      "learning_rate": 3.574821997401486e-05,
+      "loss": 0.6514,
+      "step": 8844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.439110119390619,
+      "learning_rate": 3.574445438914207e-05,
+      "loss": 0.8816,
+      "step": 8845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.488327108883691,
+      "learning_rate": 3.5740688505255474e-05,
+      "loss": 1.0693,
+      "step": 8846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6257463938094388,
+      "learning_rate": 3.573692232245991e-05,
+      "loss": 1.0059,
+      "step": 8847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1991527269660507,
+      "learning_rate": 3.573315584086015e-05,
+      "loss": 0.896,
+      "step": 8848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.023422563683367,
+      "learning_rate": 3.572938906056105e-05,
+      "loss": 1.8594,
+      "step": 8849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.021880964982029,
+      "learning_rate": 3.5725621981667404e-05,
+      "loss": 1.9512,
+      "step": 8850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7385078804386094,
+      "learning_rate": 3.5721854604284074e-05,
+      "loss": 1.4824,
+      "step": 8851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9282570297566106,
+      "learning_rate": 3.571808692851589e-05,
+      "loss": 1.1704,
+      "step": 8852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.906260130808441,
+      "learning_rate": 3.571431895446772e-05,
+      "loss": 0.9761,
+      "step": 8853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6221666030678494,
+      "learning_rate": 3.57105506822444e-05,
+      "loss": 0.9922,
+      "step": 8854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7690920242811794,
+      "learning_rate": 3.570678211195083e-05,
+      "loss": 1.4551,
+      "step": 8855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.423937573186898,
+      "learning_rate": 3.570301324369187e-05,
+      "loss": 1.0605,
+      "step": 8856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.970727874754199,
+      "learning_rate": 3.56992440775724e-05,
+      "loss": 0.6394,
+      "step": 8857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1305635899152295,
+      "learning_rate": 3.569547461369733e-05,
+      "loss": 1.2773,
+      "step": 8858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4126392601578477,
+      "learning_rate": 3.569170485217155e-05,
+      "loss": 1.1631,
+      "step": 8859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.699101711888039,
+      "learning_rate": 3.568793479309998e-05,
+      "loss": 1.4736,
+      "step": 8860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1450545987380796,
+      "learning_rate": 3.5684164436587535e-05,
+      "loss": 0.7734,
+      "step": 8861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.078495990839162,
+      "learning_rate": 3.568039378273915e-05,
+      "loss": 1.1426,
+      "step": 8862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.642411479784576,
+      "learning_rate": 3.567662283165974e-05,
+      "loss": 1.1465,
+      "step": 8863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4775944915212627,
+      "learning_rate": 3.567285158345428e-05,
+      "loss": 1.3379,
+      "step": 8864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3136302144983536,
+      "learning_rate": 3.566908003822769e-05,
+      "loss": 0.4464,
+      "step": 8865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.23252317226125,
+      "learning_rate": 3.566530819608496e-05,
+      "loss": 1.043,
+      "step": 8866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.598938222640415,
+      "learning_rate": 3.5661536057131026e-05,
+      "loss": 1.4688,
+      "step": 8867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.597484173362612,
+      "learning_rate": 3.56577636214709e-05,
+      "loss": 1.8184,
+      "step": 8868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.229677931959532,
+      "learning_rate": 3.565399088920955e-05,
+      "loss": 0.7422,
+      "step": 8869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6833587270023656,
+      "learning_rate": 3.565021786045196e-05,
+      "loss": 0.9395,
+      "step": 8870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.764905329458041,
+      "learning_rate": 3.564644453530315e-05,
+      "loss": 0.9902,
+      "step": 8871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.273019234370738,
+      "learning_rate": 3.5642670913868124e-05,
+      "loss": 1.1562,
+      "step": 8872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3230136343612937,
+      "learning_rate": 3.56388969962519e-05,
+      "loss": 1.5508,
+      "step": 8873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8757336551175565,
+      "learning_rate": 3.56351227825595e-05,
+      "loss": 1.1895,
+      "step": 8874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.163070410148289,
+      "learning_rate": 3.563134827289597e-05,
+      "loss": 1.1514,
+      "step": 8875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2403618141944657,
+      "learning_rate": 3.562757346736633e-05,
+      "loss": 0.6575,
+      "step": 8876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3146715064142587,
+      "learning_rate": 3.562379836607566e-05,
+      "loss": 1.1074,
+      "step": 8877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8978889542622834,
+      "learning_rate": 3.5620022969129e-05,
+      "loss": 0.8828,
+      "step": 8878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0329182743171184,
+      "learning_rate": 3.5616247276631435e-05,
+      "loss": 0.6138,
+      "step": 8879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7825650497273302,
+      "learning_rate": 3.561247128868802e-05,
+      "loss": 0.7256,
+      "step": 8880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.884029307371263,
+      "learning_rate": 3.560869500540385e-05,
+      "loss": 1.2754,
+      "step": 8881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.867839520815431,
+      "learning_rate": 3.560491842688402e-05,
+      "loss": 1.3047,
+      "step": 8882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.01393565273434,
+      "learning_rate": 3.560114155323362e-05,
+      "loss": 0.6709,
+      "step": 8883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6374436259407474,
+      "learning_rate": 3.559736438455777e-05,
+      "loss": 0.7852,
+      "step": 8884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.327656887606248,
+      "learning_rate": 3.559358692096159e-05,
+      "loss": 1.2764,
+      "step": 8885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.582536287302065,
+      "learning_rate": 3.558980916255019e-05,
+      "loss": 0.9106,
+      "step": 8886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.799675040371386,
+      "learning_rate": 3.558603110942872e-05,
+      "loss": 1.7695,
+      "step": 8887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.169728970749831,
+      "learning_rate": 3.558225276170231e-05,
+      "loss": 1.6914,
+      "step": 8888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.593120981951158,
+      "learning_rate": 3.557847411947611e-05,
+      "loss": 1.8555,
+      "step": 8889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8754441291423722,
+      "learning_rate": 3.557469518285528e-05,
+      "loss": 1.292,
+      "step": 8890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9444852522255447,
+      "learning_rate": 3.5570915951945e-05,
+      "loss": 1.1016,
+      "step": 8891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0894484101458186,
+      "learning_rate": 3.556713642685042e-05,
+      "loss": 0.7075,
+      "step": 8892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.171822422567798,
+      "learning_rate": 3.556335660767674e-05,
+      "loss": 0.792,
+      "step": 8893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4498487815121592,
+      "learning_rate": 3.555957649452916e-05,
+      "loss": 1.2598,
+      "step": 8894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.390745548208892,
+      "learning_rate": 3.555579608751284e-05,
+      "loss": 0.9473,
+      "step": 8895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.806451747579812,
+      "learning_rate": 3.5552015386733027e-05,
+      "loss": 1.5234,
+      "step": 8896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.338532808169527,
+      "learning_rate": 3.554823439229491e-05,
+      "loss": 0.7944,
+      "step": 8897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7627103930482244,
+      "learning_rate": 3.554445310430374e-05,
+      "loss": 1.2158,
+      "step": 8898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4982793608831355,
+      "learning_rate": 3.554067152286473e-05,
+      "loss": 1.8555,
+      "step": 8899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2919129210435734,
+      "learning_rate": 3.5536889648083114e-05,
+      "loss": 1.0205,
+      "step": 8900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8378306393673203,
+      "learning_rate": 3.5533107480064156e-05,
+      "loss": 0.9639,
+      "step": 8901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.746508180608082,
+      "learning_rate": 3.5529325018913094e-05,
+      "loss": 1.7773,
+      "step": 8902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5041023463190974,
+      "learning_rate": 3.552554226473521e-05,
+      "loss": 1.4531,
+      "step": 8903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9145371610425803,
+      "learning_rate": 3.5521759217635775e-05,
+      "loss": 1.4746,
+      "step": 8904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1304089763384924,
+      "learning_rate": 3.551797587772006e-05,
+      "loss": 0.8185,
+      "step": 8905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.451081901206012,
+      "learning_rate": 3.551419224509336e-05,
+      "loss": 0.8086,
+      "step": 8906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5390669462668676,
+      "learning_rate": 3.5510408319860966e-05,
+      "loss": 0.8623,
+      "step": 8907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.4611680141798935,
+      "learning_rate": 3.550662410212819e-05,
+      "loss": 1.6133,
+      "step": 8908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4125484365302476,
+      "learning_rate": 3.5502839592000334e-05,
+      "loss": 0.5266,
+      "step": 8909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2645302510028875,
+      "learning_rate": 3.549905478958274e-05,
+      "loss": 1.5527,
+      "step": 8910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6021553715756145,
+      "learning_rate": 3.549526969498073e-05,
+      "loss": 1.0356,
+      "step": 8911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.481495304471013,
+      "learning_rate": 3.5491484308299624e-05,
+      "loss": 1.3828,
+      "step": 8912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.80313196801508,
+      "learning_rate": 3.5487698629644784e-05,
+      "loss": 2.1484,
+      "step": 8913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.364684411450748,
+      "learning_rate": 3.548391265912156e-05,
+      "loss": 1.2891,
+      "step": 8914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.955711483652383,
+      "learning_rate": 3.5480126396835326e-05,
+      "loss": 0.9026,
+      "step": 8915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.616706836034501,
+      "learning_rate": 3.547633984289142e-05,
+      "loss": 1.0137,
+      "step": 8916
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5975025102825353,
+      "learning_rate": 3.547255299739525e-05,
+      "loss": 0.4785,
+      "step": 8917
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.449867758782596,
+      "learning_rate": 3.54687658604522e-05,
+      "loss": 0.7729,
+      "step": 8918
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2855596724226723,
+      "learning_rate": 3.5464978432167655e-05,
+      "loss": 1.2061,
+      "step": 8919
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.888932933682647,
+      "learning_rate": 3.5461190712647014e-05,
+      "loss": 1.9961,
+      "step": 8920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.407631515470851,
+      "learning_rate": 3.54574027019957e-05,
+      "loss": 0.791,
+      "step": 8921
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.124467775575348,
+      "learning_rate": 3.545361440031912e-05,
+      "loss": 1.0396,
+      "step": 8922
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0293734759062745,
+      "learning_rate": 3.544982580772272e-05,
+      "loss": 0.959,
+      "step": 8923
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.510904657781727,
+      "learning_rate": 3.5446036924311914e-05,
+      "loss": 1.2881,
+      "step": 8924
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4707253819805484,
+      "learning_rate": 3.544224775019215e-05,
+      "loss": 1.2578,
+      "step": 8925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.640837438058202,
+      "learning_rate": 3.543845828546889e-05,
+      "loss": 1.1611,
+      "step": 8926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.587805773225685,
+      "learning_rate": 3.543466853024758e-05,
+      "loss": 1.7871,
+      "step": 8927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6874842088812056,
+      "learning_rate": 3.5430878484633695e-05,
+      "loss": 1.1152,
+      "step": 8928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.498330502661805,
+      "learning_rate": 3.542708814873271e-05,
+      "loss": 1.3066,
+      "step": 8929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.480650022596282,
+      "learning_rate": 3.5423297522650095e-05,
+      "loss": 1.2363,
+      "step": 8930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4746917494120666,
+      "learning_rate": 3.541950660649136e-05,
+      "loss": 1.5293,
+      "step": 8931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2998285303747865,
+      "learning_rate": 3.5415715400362e-05,
+      "loss": 0.9375,
+      "step": 8932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.132441003763107,
+      "learning_rate": 3.5411923904367525e-05,
+      "loss": 2.0518,
+      "step": 8933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9042995618893954,
+      "learning_rate": 3.540813211861344e-05,
+      "loss": 0.9277,
+      "step": 8934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.573736435915827,
+      "learning_rate": 3.540434004320527e-05,
+      "loss": 1.4512,
+      "step": 8935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.018334866288745,
+      "learning_rate": 3.5400547678248564e-05,
+      "loss": 0.8828,
+      "step": 8936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.846446173955873,
+      "learning_rate": 3.5396755023848846e-05,
+      "loss": 1.4551,
+      "step": 8937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0082615450405292,
+      "learning_rate": 3.539296208011167e-05,
+      "loss": 0.6541,
+      "step": 8938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.06464863611974,
+      "learning_rate": 3.538916884714258e-05,
+      "loss": 1.2324,
+      "step": 8939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6356686101287865,
+      "learning_rate": 3.5385375325047166e-05,
+      "loss": 0.9331,
+      "step": 8940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8472456354765447,
+      "learning_rate": 3.5381581513930974e-05,
+      "loss": 1.0337,
+      "step": 8941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.141439575247748,
+      "learning_rate": 3.53777874138996e-05,
+      "loss": 1.1523,
+      "step": 8942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7206674588529514,
+      "learning_rate": 3.537399302505862e-05,
+      "loss": 0.9316,
+      "step": 8943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.941853849612851,
+      "learning_rate": 3.537019834751364e-05,
+      "loss": 1.1807,
+      "step": 8944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2652358148356133,
+      "learning_rate": 3.536640338137027e-05,
+      "loss": 0.7842,
+      "step": 8945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.507410432465594,
+      "learning_rate": 3.53626081267341e-05,
+      "loss": 1.1172,
+      "step": 8946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.844840710409379,
+      "learning_rate": 3.5358812583710784e-05,
+      "loss": 0.918,
+      "step": 8947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.019125016504155,
+      "learning_rate": 3.535501675240591e-05,
+      "loss": 1.4102,
+      "step": 8948
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2826148691292314,
+      "learning_rate": 3.5351220632925145e-05,
+      "loss": 1.2188,
+      "step": 8949
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3988942261275703,
+      "learning_rate": 3.534742422537413e-05,
+      "loss": 2.0898,
+      "step": 8950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2239831808401043,
+      "learning_rate": 3.5343627529858504e-05,
+      "loss": 0.8293,
+      "step": 8951
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8260227734084664,
+      "learning_rate": 3.533983054648393e-05,
+      "loss": 1.2734,
+      "step": 8952
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.198402673245369,
+      "learning_rate": 3.5336033275356085e-05,
+      "loss": 1.2148,
+      "step": 8953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.010007736693265,
+      "learning_rate": 3.533223571658064e-05,
+      "loss": 0.7422,
+      "step": 8954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.139913844575765,
+      "learning_rate": 3.532843787026329e-05,
+      "loss": 0.8018,
+      "step": 8955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3101137582420668,
+      "learning_rate": 3.5324639736509714e-05,
+      "loss": 1.0879,
+      "step": 8956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.611355690224655,
+      "learning_rate": 3.532084131542562e-05,
+      "loss": 1.3047,
+      "step": 8957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2933458476316573,
+      "learning_rate": 3.53170426071167e-05,
+      "loss": 1.3018,
+      "step": 8958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.672035703232074,
+      "learning_rate": 3.53132436116887e-05,
+      "loss": 0.6807,
+      "step": 8959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.012600307452673,
+      "learning_rate": 3.530944432924732e-05,
+      "loss": 1.375,
+      "step": 8960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.024087866749771,
+      "learning_rate": 3.530564475989831e-05,
+      "loss": 0.9492,
+      "step": 8961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3730966068284234,
+      "learning_rate": 3.5301844903747395e-05,
+      "loss": 1.2207,
+      "step": 8962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.378237074267597,
+      "learning_rate": 3.529804476090033e-05,
+      "loss": 1.6992,
+      "step": 8963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.170388761442615,
+      "learning_rate": 3.5294244331462876e-05,
+      "loss": 1.1523,
+      "step": 8964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9912598768135754,
+      "learning_rate": 3.529044361554079e-05,
+      "loss": 1.0996,
+      "step": 8965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.165770938188268,
+      "learning_rate": 3.528664261323985e-05,
+      "loss": 0.8984,
+      "step": 8966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.199244931266664,
+      "learning_rate": 3.528284132466583e-05,
+      "loss": 1.123,
+      "step": 8967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4529563700065107,
+      "learning_rate": 3.527903974992452e-05,
+      "loss": 1.3242,
+      "step": 8968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.464905654927854,
+      "learning_rate": 3.527523788912172e-05,
+      "loss": 1.2275,
+      "step": 8969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.364537231623276,
+      "learning_rate": 3.527143574236324e-05,
+      "loss": 1.1562,
+      "step": 8970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0825776557391236,
+      "learning_rate": 3.526763330975487e-05,
+      "loss": 0.9297,
+      "step": 8971
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.813739368473088,
+      "learning_rate": 3.526383059140246e-05,
+      "loss": 1.8711,
+      "step": 8972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8329903077061753,
+      "learning_rate": 3.526002758741182e-05,
+      "loss": 0.8838,
+      "step": 8973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.649890338204825,
+      "learning_rate": 3.5256224297888784e-05,
+      "loss": 0.4841,
+      "step": 8974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.6698801615327445,
+      "learning_rate": 3.525242072293921e-05,
+      "loss": 1.2334,
+      "step": 8975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.309355789125314,
+      "learning_rate": 3.524861686266893e-05,
+      "loss": 1.002,
+      "step": 8976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1390574932642452,
+      "learning_rate": 3.524481271718382e-05,
+      "loss": 0.3672,
+      "step": 8977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6631648548372602,
+      "learning_rate": 3.524100828658974e-05,
+      "loss": 0.4747,
+      "step": 8978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.31780119863478,
+      "learning_rate": 3.523720357099257e-05,
+      "loss": 1.1787,
+      "step": 8979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.661673281326072,
+      "learning_rate": 3.523339857049819e-05,
+      "loss": 0.8594,
+      "step": 8980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5198303840692646,
+      "learning_rate": 3.52295932852125e-05,
+      "loss": 1.0312,
+      "step": 8981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.9194814523494,
+      "learning_rate": 3.5225787715241385e-05,
+      "loss": 1.0391,
+      "step": 8982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.845493042425541,
+      "learning_rate": 3.5221981860690775e-05,
+      "loss": 2.2832,
+      "step": 8983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5495378112802056,
+      "learning_rate": 3.5218175721666554e-05,
+      "loss": 0.7126,
+      "step": 8984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.099511007165984,
+      "learning_rate": 3.521436929827467e-05,
+      "loss": 1.0742,
+      "step": 8985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6429895408997512,
+      "learning_rate": 3.521056259062105e-05,
+      "loss": 0.8057,
+      "step": 8986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7502498513207745,
+      "learning_rate": 3.520675559881163e-05,
+      "loss": 1.1074,
+      "step": 8987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1763326956752693,
+      "learning_rate": 3.520294832295236e-05,
+      "loss": 1.498,
+      "step": 8988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.561680615876342,
+      "learning_rate": 3.519914076314918e-05,
+      "loss": 1.0029,
+      "step": 8989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.226002833649922,
+      "learning_rate": 3.519533291950807e-05,
+      "loss": 1.2285,
+      "step": 8990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5308540594881994,
+      "learning_rate": 3.5191524792134996e-05,
+      "loss": 1.2539,
+      "step": 8991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.248842384232138,
+      "learning_rate": 3.518771638113593e-05,
+      "loss": 1.3359,
+      "step": 8992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.404261175862101,
+      "learning_rate": 3.518390768661688e-05,
+      "loss": 1.3379,
+      "step": 8993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3647157965260526,
+      "learning_rate": 3.518009870868381e-05,
+      "loss": 1.3984,
+      "step": 8994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3819651613610437,
+      "learning_rate": 3.517628944744274e-05,
+      "loss": 0.7461,
+      "step": 8995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.101620758263119,
+      "learning_rate": 3.5172479902999685e-05,
+      "loss": 0.8743,
+      "step": 8996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4783164952143744,
+      "learning_rate": 3.5168670075460646e-05,
+      "loss": 1.2305,
+      "step": 8997
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.709678775428837,
+      "learning_rate": 3.516485996493166e-05,
+      "loss": 0.71,
+      "step": 8998
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.227218745924331,
+      "learning_rate": 3.516104957151877e-05,
+      "loss": 1.001,
+      "step": 8999
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.226385542862756,
+      "learning_rate": 3.5157238895328e-05,
+      "loss": 1.3652,
+      "step": 9000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0316191117831486,
+      "learning_rate": 3.51534279364654e-05,
+      "loss": 1.2378,
+      "step": 9001
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.075728256585957,
+      "learning_rate": 3.514961669503705e-05,
+      "loss": 1.2139,
+      "step": 9002
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.013090500675144,
+      "learning_rate": 3.514580517114898e-05,
+      "loss": 0.9844,
+      "step": 9003
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.519840953443622,
+      "learning_rate": 3.514199336490729e-05,
+      "loss": 2.5332,
+      "step": 9004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1368993419759987,
+      "learning_rate": 3.513818127641805e-05,
+      "loss": 1.0684,
+      "step": 9005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2835538496397296,
+      "learning_rate": 3.513436890578736e-05,
+      "loss": 0.9512,
+      "step": 9006
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1007278818394877,
+      "learning_rate": 3.5130556253121314e-05,
+      "loss": 1.0195,
+      "step": 9007
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.074046777361492,
+      "learning_rate": 3.5126743318526e-05,
+      "loss": 0.7441,
+      "step": 9008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5101412124729507,
+      "learning_rate": 3.512293010210754e-05,
+      "loss": 0.73,
+      "step": 9009
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7739277712688017,
+      "learning_rate": 3.511911660397207e-05,
+      "loss": 0.9785,
+      "step": 9010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7439084643786265,
+      "learning_rate": 3.5115302824225695e-05,
+      "loss": 1.0947,
+      "step": 9011
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.367227292987386,
+      "learning_rate": 3.511148876297457e-05,
+      "loss": 1.0898,
+      "step": 9012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.461600949911009,
+      "learning_rate": 3.510767442032482e-05,
+      "loss": 1.043,
+      "step": 9013
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8070711935138593,
+      "learning_rate": 3.51038597963826e-05,
+      "loss": 1.2773,
+      "step": 9014
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3664118437072794,
+      "learning_rate": 3.5100044891254084e-05,
+      "loss": 1.2148,
+      "step": 9015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7782673542447456,
+      "learning_rate": 3.509622970504543e-05,
+      "loss": 0.5732,
+      "step": 9016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5951682867163441,
+      "learning_rate": 3.509241423786281e-05,
+      "loss": 0.6009,
+      "step": 9017
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.147434887247814,
+      "learning_rate": 3.508859848981241e-05,
+      "loss": 1.6548,
+      "step": 9018
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.091515672661928,
+      "learning_rate": 3.508478246100042e-05,
+      "loss": 1.4277,
+      "step": 9019
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2659188967851125,
+      "learning_rate": 3.508096615153305e-05,
+      "loss": 1.1025,
+      "step": 9020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1932725218291926,
+      "learning_rate": 3.507714956151649e-05,
+      "loss": 1.0596,
+      "step": 9021
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2108746307384815,
+      "learning_rate": 3.5073332691056956e-05,
+      "loss": 1.0957,
+      "step": 9022
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.516348916075715,
+      "learning_rate": 3.5069515540260675e-05,
+      "loss": 1.2988,
+      "step": 9023
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.551391341664448,
+      "learning_rate": 3.506569810923388e-05,
+      "loss": 0.9521,
+      "step": 9024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.107061403007155,
+      "learning_rate": 3.50618803980828e-05,
+      "loss": 1.3613,
+      "step": 9025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3176744460327026,
+      "learning_rate": 3.505806240691369e-05,
+      "loss": 1.166,
+      "step": 9026
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.552712562885843,
+      "learning_rate": 3.505424413583279e-05,
+      "loss": 1.123,
+      "step": 9027
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.017014355613664,
+      "learning_rate": 3.505042558494637e-05,
+      "loss": 0.9326,
+      "step": 9028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.93179529093302,
+      "learning_rate": 3.504660675436071e-05,
+      "loss": 1.1582,
+      "step": 9029
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4314356630328495,
+      "learning_rate": 3.504278764418206e-05,
+      "loss": 1.1895,
+      "step": 9030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5468317104840366,
+      "learning_rate": 3.503896825451673e-05,
+      "loss": 2.0078,
+      "step": 9031
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0456319542113,
+      "learning_rate": 3.5035148585470986e-05,
+      "loss": 1.6523,
+      "step": 9032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.571111583850714,
+      "learning_rate": 3.5031328637151154e-05,
+      "loss": 1.2754,
+      "step": 9033
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.8557608778570645,
+      "learning_rate": 3.5027508409663526e-05,
+      "loss": 1.5977,
+      "step": 9034
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4062045033067476,
+      "learning_rate": 3.502368790311442e-05,
+      "loss": 1.3174,
+      "step": 9035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7999355683406595,
+      "learning_rate": 3.501986711761016e-05,
+      "loss": 1.3105,
+      "step": 9036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.8446443777989385,
+      "learning_rate": 3.501604605325708e-05,
+      "loss": 1.2207,
+      "step": 9037
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.054363964992372,
+      "learning_rate": 3.501222471016152e-05,
+      "loss": 0.7363,
+      "step": 9038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9990375706518284,
+      "learning_rate": 3.5008403088429816e-05,
+      "loss": 1.2354,
+      "step": 9039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6694915134418644,
+      "learning_rate": 3.500458118816833e-05,
+      "loss": 1.1211,
+      "step": 9040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7896339835444355,
+      "learning_rate": 3.5000759009483415e-05,
+      "loss": 1.1221,
+      "step": 9041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.305527902811807,
+      "learning_rate": 3.499693655248146e-05,
+      "loss": 1.293,
+      "step": 9042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1857337360424913,
+      "learning_rate": 3.499311381726882e-05,
+      "loss": 0.4438,
+      "step": 9043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0098141436421635,
+      "learning_rate": 3.49892908039519e-05,
+      "loss": 1.332,
+      "step": 9044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.206704593847141,
+      "learning_rate": 3.4985467512637074e-05,
+      "loss": 1.6973,
+      "step": 9045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.029937774724288,
+      "learning_rate": 3.498164394343076e-05,
+      "loss": 1.1289,
+      "step": 9046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1440550639060136,
+      "learning_rate": 3.4977820096439353e-05,
+      "loss": 1.2461,
+      "step": 9047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4001420018784114,
+      "learning_rate": 3.497399597176928e-05,
+      "loss": 1.1855,
+      "step": 9048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.951457882088698,
+      "learning_rate": 3.4970171569526956e-05,
+      "loss": 0.8945,
+      "step": 9049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.288353194919011,
+      "learning_rate": 3.496634688981882e-05,
+      "loss": 1.2349,
+      "step": 9050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.875666997080953,
+      "learning_rate": 3.49625219327513e-05,
+      "loss": 1.5898,
+      "step": 9051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4505699992789065,
+      "learning_rate": 3.495869669843086e-05,
+      "loss": 0.8447,
+      "step": 9052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4036731546212406,
+      "learning_rate": 3.495487118696394e-05,
+      "loss": 1.334,
+      "step": 9053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.299698474864698,
+      "learning_rate": 3.4951045398457005e-05,
+      "loss": 0.855,
+      "step": 9054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.469744164913723,
+      "learning_rate": 3.4947219333016534e-05,
+      "loss": 0.8096,
+      "step": 9055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.427040277614503,
+      "learning_rate": 3.494339299074899e-05,
+      "loss": 0.7808,
+      "step": 9056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6763429121399436,
+      "learning_rate": 3.493956637176088e-05,
+      "loss": 1.1572,
+      "step": 9057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.178522379678149,
+      "learning_rate": 3.493573947615867e-05,
+      "loss": 0.4521,
+      "step": 9058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.449842650362599,
+      "learning_rate": 3.493191230404888e-05,
+      "loss": 0.8994,
+      "step": 9059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.899803885395195,
+      "learning_rate": 3.492808485553801e-05,
+      "loss": 1.3896,
+      "step": 9060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.678211718302787,
+      "learning_rate": 3.49242571307326e-05,
+      "loss": 1.4863,
+      "step": 9061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.493756891303233,
+      "learning_rate": 3.4920429129739126e-05,
+      "loss": 1.1436,
+      "step": 9062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.444937387768579,
+      "learning_rate": 3.4916600852664175e-05,
+      "loss": 1.1875,
+      "step": 9063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.900872886004633,
+      "learning_rate": 3.491277229961424e-05,
+      "loss": 0.7112,
+      "step": 9064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.017122045327528,
+      "learning_rate": 3.4908943470695895e-05,
+      "loss": 0.623,
+      "step": 9065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1610791116654324,
+      "learning_rate": 3.490511436601568e-05,
+      "loss": 1.1777,
+      "step": 9066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6606793442477232,
+      "learning_rate": 3.490128498568017e-05,
+      "loss": 1.0693,
+      "step": 9067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.533150110094999,
+      "learning_rate": 3.489745532979593e-05,
+      "loss": 1.4023,
+      "step": 9068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.082908669860205,
+      "learning_rate": 3.4893625398469536e-05,
+      "loss": 0.8267,
+      "step": 9069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.594343235424748,
+      "learning_rate": 3.488979519180758e-05,
+      "loss": 1.2742,
+      "step": 9070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.2947388789384675,
+      "learning_rate": 3.488596470991664e-05,
+      "loss": 0.9785,
+      "step": 9071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.45853687241193,
+      "learning_rate": 3.488213395290334e-05,
+      "loss": 0.6752,
+      "step": 9072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.742101024360841,
+      "learning_rate": 3.4878302920874264e-05,
+      "loss": 1.0112,
+      "step": 9073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.286292425079851,
+      "learning_rate": 3.487447161393605e-05,
+      "loss": 1.2559,
+      "step": 9074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8983629736354484,
+      "learning_rate": 3.4870640032195296e-05,
+      "loss": 1.209,
+      "step": 9075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8319583436197573,
+      "learning_rate": 3.486680817575866e-05,
+      "loss": 0.8613,
+      "step": 9076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.619656255317978,
+      "learning_rate": 3.486297604473277e-05,
+      "loss": 0.3315,
+      "step": 9077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.379616817594354,
+      "learning_rate": 3.485914363922426e-05,
+      "loss": 1.1719,
+      "step": 9078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.400409976700391,
+      "learning_rate": 3.48553109593398e-05,
+      "loss": 1.2373,
+      "step": 9079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3439111403381427,
+      "learning_rate": 3.4851478005186054e-05,
+      "loss": 0.9434,
+      "step": 9080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.037974812637399,
+      "learning_rate": 3.484764477686968e-05,
+      "loss": 1.1465,
+      "step": 9081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.26319773550408,
+      "learning_rate": 3.484381127449737e-05,
+      "loss": 1.199,
+      "step": 9082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5549384730684506,
+      "learning_rate": 3.483997749817579e-05,
+      "loss": 1.3789,
+      "step": 9083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0554348160749893,
+      "learning_rate": 3.4836143448011644e-05,
+      "loss": 1.0371,
+      "step": 9084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.517121053774138,
+      "learning_rate": 3.483230912411163e-05,
+      "loss": 1.2148,
+      "step": 9085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.914373196966973,
+      "learning_rate": 3.482847452658246e-05,
+      "loss": 1.4316,
+      "step": 9086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.34640252041831,
+      "learning_rate": 3.4824639655530845e-05,
+      "loss": 0.9233,
+      "step": 9087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.295216685134275,
+      "learning_rate": 3.4820804511063496e-05,
+      "loss": 1.4609,
+      "step": 9088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.407807936028575,
+      "learning_rate": 3.481696909328718e-05,
+      "loss": 1.207,
+      "step": 9089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1905172690381667,
+      "learning_rate": 3.481313340230859e-05,
+      "loss": 1.0771,
+      "step": 9090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.614517170548936,
+      "learning_rate": 3.48092974382345e-05,
+      "loss": 0.8564,
+      "step": 9091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0420036303456532,
+      "learning_rate": 3.4805461201171644e-05,
+      "loss": 1.2637,
+      "step": 9092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4774826698967387,
+      "learning_rate": 3.480162469122681e-05,
+      "loss": 0.9385,
+      "step": 9093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.936085769357335,
+      "learning_rate": 3.479778790850675e-05,
+      "loss": 0.5823,
+      "step": 9094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0215703531308207,
+      "learning_rate": 3.479395085311824e-05,
+      "loss": 1.0269,
+      "step": 9095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.47503894428478,
+      "learning_rate": 3.479011352516806e-05,
+      "loss": 1.1274,
+      "step": 9096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.25174507708608,
+      "learning_rate": 3.4786275924763e-05,
+      "loss": 0.8193,
+      "step": 9097
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.363551167074495,
+      "learning_rate": 3.478243805200987e-05,
+      "loss": 0.9775,
+      "step": 9098
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9335109546664486,
+      "learning_rate": 3.477859990701548e-05,
+      "loss": 1.0581,
+      "step": 9099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.46325197957982,
+      "learning_rate": 3.4774761489886626e-05,
+      "loss": 0.6721,
+      "step": 9100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6812093062167435,
+      "learning_rate": 3.4770922800730145e-05,
+      "loss": 1.3574,
+      "step": 9101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2622439708530306,
+      "learning_rate": 3.476708383965286e-05,
+      "loss": 0.811,
+      "step": 9102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.823458874045757,
+      "learning_rate": 3.47632446067616e-05,
+      "loss": 0.7944,
+      "step": 9103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4803921424870454,
+      "learning_rate": 3.475940510216323e-05,
+      "loss": 1.3242,
+      "step": 9104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.152722761613717,
+      "learning_rate": 3.475556532596458e-05,
+      "loss": 1.7754,
+      "step": 9105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0720039348378414,
+      "learning_rate": 3.475172527827252e-05,
+      "loss": 1.0498,
+      "step": 9106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.744782266232766,
+      "learning_rate": 3.474788495919391e-05,
+      "loss": 0.7373,
+      "step": 9107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.995260662803841,
+      "learning_rate": 3.4744044368835643e-05,
+      "loss": 1.2363,
+      "step": 9108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3383919059756892,
+      "learning_rate": 3.474020350730458e-05,
+      "loss": 1.291,
+      "step": 9109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5522019546863612,
+      "learning_rate": 3.4736362374707614e-05,
+      "loss": 1.2148,
+      "step": 9110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0799886048444973,
+      "learning_rate": 3.473252097115165e-05,
+      "loss": 0.8486,
+      "step": 9111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7267531500137885,
+      "learning_rate": 3.47286792967436e-05,
+      "loss": 1.5527,
+      "step": 9112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.470685866906463,
+      "learning_rate": 3.472483735159036e-05,
+      "loss": 1.9434,
+      "step": 9113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.612191370328001,
+      "learning_rate": 3.472099513579885e-05,
+      "loss": 1.3777,
+      "step": 9114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6325301522081905,
+      "learning_rate": 3.4717152649476e-05,
+      "loss": 0.9087,
+      "step": 9115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.081157823315416,
+      "learning_rate": 3.471330989272876e-05,
+      "loss": 1.1436,
+      "step": 9116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9052456997100253,
+      "learning_rate": 3.470946686566405e-05,
+      "loss": 0.9521,
+      "step": 9117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6217232188507213,
+      "learning_rate": 3.470562356838883e-05,
+      "loss": 1.3691,
+      "step": 9118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.131979084320248,
+      "learning_rate": 3.470178000101006e-05,
+      "loss": 1.4688,
+      "step": 9119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.72732979107893,
+      "learning_rate": 3.46979361636347e-05,
+      "loss": 1.1299,
+      "step": 9120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.250091327359161,
+      "learning_rate": 3.469409205636973e-05,
+      "loss": 0.9619,
+      "step": 9121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1759645669029717,
+      "learning_rate": 3.469024767932212e-05,
+      "loss": 1.0332,
+      "step": 9122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.094210364010117,
+      "learning_rate": 3.468640303259887e-05,
+      "loss": 1.4092,
+      "step": 9123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.017971094709725,
+      "learning_rate": 3.468255811630696e-05,
+      "loss": 1.2812,
+      "step": 9124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2182490898241376,
+      "learning_rate": 3.4678712930553394e-05,
+      "loss": 1.8477,
+      "step": 9125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.013060868008429,
+      "learning_rate": 3.4674867475445196e-05,
+      "loss": 1.1484,
+      "step": 9126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6202579535833204,
+      "learning_rate": 3.4671021751089374e-05,
+      "loss": 1.2207,
+      "step": 9127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2918772398410043,
+      "learning_rate": 3.4667175757592953e-05,
+      "loss": 0.8647,
+      "step": 9128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.769182148225365,
+      "learning_rate": 3.4663329495062965e-05,
+      "loss": 1.4961,
+      "step": 9129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.247072453913477,
+      "learning_rate": 3.4659482963606446e-05,
+      "loss": 1.4727,
+      "step": 9130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.525370138251928,
+      "learning_rate": 3.465563616333046e-05,
+      "loss": 0.9277,
+      "step": 9131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.05791761959672,
+      "learning_rate": 3.4651789094342044e-05,
+      "loss": 0.8384,
+      "step": 9132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9211983126636882,
+      "learning_rate": 3.464794175674827e-05,
+      "loss": 1.1484,
+      "step": 9133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.679155124463442,
+      "learning_rate": 3.464409415065619e-05,
+      "loss": 1.2754,
+      "step": 9134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5679641709665497,
+      "learning_rate": 3.464024627617292e-05,
+      "loss": 0.835,
+      "step": 9135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.061941610553439,
+      "learning_rate": 3.46363981334055e-05,
+      "loss": 0.7268,
+      "step": 9136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9590568491041425,
+      "learning_rate": 3.463254972246106e-05,
+      "loss": 0.6772,
+      "step": 9137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8281040736035172,
+      "learning_rate": 3.462870104344666e-05,
+      "loss": 1.1387,
+      "step": 9138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.911359336047983,
+      "learning_rate": 3.462485209646945e-05,
+      "loss": 0.6072,
+      "step": 9139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9802039456573899,
+      "learning_rate": 3.4621002881636525e-05,
+      "loss": 1.1357,
+      "step": 9140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.138874144078533,
+      "learning_rate": 3.461715339905499e-05,
+      "loss": 1.6309,
+      "step": 9141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.298203363562619,
+      "learning_rate": 3.4613303648832006e-05,
+      "loss": 1.584,
+      "step": 9142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7894644995162206,
+      "learning_rate": 3.460945363107469e-05,
+      "loss": 1.1758,
+      "step": 9143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.051108212112688,
+      "learning_rate": 3.460560334589019e-05,
+      "loss": 1.6113,
+      "step": 9144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2259926585430265,
+      "learning_rate": 3.4601752793385655e-05,
+      "loss": 1.3242,
+      "step": 9145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4577746214375953,
+      "learning_rate": 3.4597901973668255e-05,
+      "loss": 0.9214,
+      "step": 9146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3192142237800484,
+      "learning_rate": 3.4594050886845154e-05,
+      "loss": 0.5869,
+      "step": 9147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5725345833601385,
+      "learning_rate": 3.459019953302351e-05,
+      "loss": 0.998,
+      "step": 9148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.021633822611002,
+      "learning_rate": 3.4586347912310524e-05,
+      "loss": 0.8115,
+      "step": 9149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.951395259078761,
+      "learning_rate": 3.458249602481338e-05,
+      "loss": 1.3633,
+      "step": 9150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.018076264988917,
+      "learning_rate": 3.457864387063928e-05,
+      "loss": 1.6582,
+      "step": 9151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.072379234429931,
+      "learning_rate": 3.4574791449895406e-05,
+      "loss": 1.2275,
+      "step": 9152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7217937400416803,
+      "learning_rate": 3.457093876268899e-05,
+      "loss": 1.3457,
+      "step": 9153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.140869137364662,
+      "learning_rate": 3.456708580912725e-05,
+      "loss": 1.7202,
+      "step": 9154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.438893726675246,
+      "learning_rate": 3.4563232589317395e-05,
+      "loss": 1.3916,
+      "step": 9155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.571907938257005,
+      "learning_rate": 3.4559379103366686e-05,
+      "loss": 1.3809,
+      "step": 9156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.404833437841111,
+      "learning_rate": 3.4555525351382334e-05,
+      "loss": 0.9023,
+      "step": 9157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8351178626801543,
+      "learning_rate": 3.45516713334716e-05,
+      "loss": 1.2227,
+      "step": 9158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.265785644978563,
+      "learning_rate": 3.4547817049741754e-05,
+      "loss": 0.8325,
+      "step": 9159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.56394471171693,
+      "learning_rate": 3.454396250030004e-05,
+      "loss": 0.9844,
+      "step": 9160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1900836720748984,
+      "learning_rate": 3.454010768525373e-05,
+      "loss": 0.4248,
+      "step": 9161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.443700484532724,
+      "learning_rate": 3.4536252604710115e-05,
+      "loss": 0.781,
+      "step": 9162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7074632945594175,
+      "learning_rate": 3.453239725877647e-05,
+      "loss": 0.8242,
+      "step": 9163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.9351230079610735,
+      "learning_rate": 3.452854164756009e-05,
+      "loss": 1.1973,
+      "step": 9164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4747040812518177,
+      "learning_rate": 3.4524685771168276e-05,
+      "loss": 0.8916,
+      "step": 9165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7501309689857365,
+      "learning_rate": 3.452082962970833e-05,
+      "loss": 1.1973,
+      "step": 9166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0495633241726643,
+      "learning_rate": 3.4516973223287576e-05,
+      "loss": 0.6953,
+      "step": 9167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.595758822049618,
+      "learning_rate": 3.451311655201333e-05,
+      "loss": 1.2207,
+      "step": 9168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.448695952912899,
+      "learning_rate": 3.450925961599292e-05,
+      "loss": 1.2334,
+      "step": 9169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.575165418990408,
+      "learning_rate": 3.450540241533369e-05,
+      "loss": 1.4023,
+      "step": 9170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4542873721934058,
+      "learning_rate": 3.4501544950142975e-05,
+      "loss": 1.1797,
+      "step": 9171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.475459867197945,
+      "learning_rate": 3.449768722052814e-05,
+      "loss": 0.7422,
+      "step": 9172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2061388760427034,
+      "learning_rate": 3.449382922659653e-05,
+      "loss": 1.335,
+      "step": 9173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.289252992587015,
+      "learning_rate": 3.448997096845553e-05,
+      "loss": 1.1133,
+      "step": 9174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.913681179132248,
+      "learning_rate": 3.448611244621248e-05,
+      "loss": 0.863,
+      "step": 9175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1061766771224577,
+      "learning_rate": 3.448225365997479e-05,
+      "loss": 1.3379,
+      "step": 9176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8526172934111678,
+      "learning_rate": 3.447839460984985e-05,
+      "loss": 0.9199,
+      "step": 9177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.876085013908823,
+      "learning_rate": 3.4474535295945044e-05,
+      "loss": 1.5049,
+      "step": 9178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4194317707012556,
+      "learning_rate": 3.4470675718367775e-05,
+      "loss": 0.4382,
+      "step": 9179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.324417927767433,
+      "learning_rate": 3.446681587722546e-05,
+      "loss": 1.291,
+      "step": 9180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.40860809642912,
+      "learning_rate": 3.44629557726255e-05,
+      "loss": 1.375,
+      "step": 9181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.820340058675518,
+      "learning_rate": 3.4459095404675354e-05,
+      "loss": 0.8018,
+      "step": 9182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.325712938893263,
+      "learning_rate": 3.4455234773482415e-05,
+      "loss": 1.3311,
+      "step": 9183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.403043865431186,
+      "learning_rate": 3.445137387915415e-05,
+      "loss": 1.0615,
+      "step": 9184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.605602163762385,
+      "learning_rate": 3.444751272179799e-05,
+      "loss": 1.0259,
+      "step": 9185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.857556984382259,
+      "learning_rate": 3.4443651301521404e-05,
+      "loss": 0.7402,
+      "step": 9186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7659757458700454,
+      "learning_rate": 3.443978961843185e-05,
+      "loss": 0.7227,
+      "step": 9187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.941280286303095,
+      "learning_rate": 3.443592767263679e-05,
+      "loss": 1.6699,
+      "step": 9188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0784348075373975,
+      "learning_rate": 3.4432065464243704e-05,
+      "loss": 1.2793,
+      "step": 9189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0141063285555916,
+      "learning_rate": 3.442820299336007e-05,
+      "loss": 0.9463,
+      "step": 9190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.05509620613787,
+      "learning_rate": 3.4424340260093404e-05,
+      "loss": 1.5811,
+      "step": 9191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.5451593996807045,
+      "learning_rate": 3.442047726455117e-05,
+      "loss": 1.1553,
+      "step": 9192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.601819275717192,
+      "learning_rate": 3.44166140068409e-05,
+      "loss": 1.582,
+      "step": 9193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.492274752889045,
+      "learning_rate": 3.4412750487070074e-05,
+      "loss": 0.9844,
+      "step": 9194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.945346012481343,
+      "learning_rate": 3.4408886705346246e-05,
+      "loss": 1.3203,
+      "step": 9195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.634087372675216,
+      "learning_rate": 3.4405022661776935e-05,
+      "loss": 1.4844,
+      "step": 9196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.464679707375284,
+      "learning_rate": 3.440115835646968e-05,
+      "loss": 0.9844,
+      "step": 9197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.24129359273647,
+      "learning_rate": 3.439729378953199e-05,
+      "loss": 0.8867,
+      "step": 9198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9619496874938873,
+      "learning_rate": 3.439342896107146e-05,
+      "loss": 0.7178,
+      "step": 9199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2653623231346733,
+      "learning_rate": 3.438956387119562e-05,
+      "loss": 1.2969,
+      "step": 9200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.713785653713485,
+      "learning_rate": 3.438569852001204e-05,
+      "loss": 1.5273,
+      "step": 9201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5714470582630526,
+      "learning_rate": 3.4381832907628286e-05,
+      "loss": 1.3242,
+      "step": 9202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.82121342775153,
+      "learning_rate": 3.4377967034151947e-05,
+      "loss": 1.0098,
+      "step": 9203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9174254477280765,
+      "learning_rate": 3.43741008996906e-05,
+      "loss": 1.2051,
+      "step": 9204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.399166331934356,
+      "learning_rate": 3.437023450435185e-05,
+      "loss": 1.5234,
+      "step": 9205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.918847785308242,
+      "learning_rate": 3.436636784824328e-05,
+      "loss": 0.6738,
+      "step": 9206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.752700265557834,
+      "learning_rate": 3.436250093147251e-05,
+      "loss": 0.9824,
+      "step": 9207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7949849993963447,
+      "learning_rate": 3.435863375414715e-05,
+      "loss": 1.5254,
+      "step": 9208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.085942345126513,
+      "learning_rate": 3.435476631637482e-05,
+      "loss": 1.6621,
+      "step": 9209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.821908885045711,
+      "learning_rate": 3.4350898618263156e-05,
+      "loss": 1.5186,
+      "step": 9210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5457311764796358,
+      "learning_rate": 3.434703065991979e-05,
+      "loss": 1.1362,
+      "step": 9211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7963388721245943,
+      "learning_rate": 3.434316244145236e-05,
+      "loss": 0.5767,
+      "step": 9212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.011656291695773,
+      "learning_rate": 3.433929396296852e-05,
+      "loss": 0.7495,
+      "step": 9213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8080755360572591,
+      "learning_rate": 3.433542522457594e-05,
+      "loss": 0.6003,
+      "step": 9214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2720495227802724,
+      "learning_rate": 3.433155622638227e-05,
+      "loss": 1.0566,
+      "step": 9215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.698181715945066,
+      "learning_rate": 3.43276869684952e-05,
+      "loss": 0.7783,
+      "step": 9216
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.235461524401907,
+      "learning_rate": 3.432381745102238e-05,
+      "loss": 1.1011,
+      "step": 9217
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.695930741764723,
+      "learning_rate": 3.4319947674071535e-05,
+      "loss": 0.9448,
+      "step": 9218
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9312292492748155,
+      "learning_rate": 3.4316077637750336e-05,
+      "loss": 0.9504,
+      "step": 9219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2657128678745018,
+      "learning_rate": 3.4312207342166487e-05,
+      "loss": 0.5355,
+      "step": 9220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9476755715350382,
+      "learning_rate": 3.430833678742769e-05,
+      "loss": 1.0098,
+      "step": 9221
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.580295827894205,
+      "learning_rate": 3.430446597364168e-05,
+      "loss": 0.9414,
+      "step": 9222
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3668770665234535,
+      "learning_rate": 3.430059490091617e-05,
+      "loss": 1.2031,
+      "step": 9223
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9730820602273467,
+      "learning_rate": 3.429672356935889e-05,
+      "loss": 0.8496,
+      "step": 9224
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.018459403297755,
+      "learning_rate": 3.429285197907757e-05,
+      "loss": 0.8945,
+      "step": 9225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1918970335316987,
+      "learning_rate": 3.428898013017997e-05,
+      "loss": 1.4062,
+      "step": 9226
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.05812522000828,
+      "learning_rate": 3.4285108022773826e-05,
+      "loss": 1.1426,
+      "step": 9227
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.4285926071223125,
+      "learning_rate": 3.4281235656966915e-05,
+      "loss": 1.2988,
+      "step": 9228
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.518615177606995,
+      "learning_rate": 3.4277363032866996e-05,
+      "loss": 1.1035,
+      "step": 9229
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.549410140917313,
+      "learning_rate": 3.4273490150581834e-05,
+      "loss": 2.3262,
+      "step": 9230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.125927901790504,
+      "learning_rate": 3.426961701021921e-05,
+      "loss": 1.3906,
+      "step": 9231
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4036924964505495,
+      "learning_rate": 3.426574361188692e-05,
+      "loss": 1.0723,
+      "step": 9232
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.390989783287097,
+      "learning_rate": 3.426186995569277e-05,
+      "loss": 1.0732,
+      "step": 9233
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9245372718132723,
+      "learning_rate": 3.425799604174455e-05,
+      "loss": 1.1475,
+      "step": 9234
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.901199780163952,
+      "learning_rate": 3.425412187015006e-05,
+      "loss": 0.96,
+      "step": 9235
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1260022436508725,
+      "learning_rate": 3.4250247441017124e-05,
+      "loss": 1.2246,
+      "step": 9236
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.079712229410064,
+      "learning_rate": 3.424637275445357e-05,
+      "loss": 0.4993,
+      "step": 9237
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.047873244577031,
+      "learning_rate": 3.424249781056723e-05,
+      "loss": 1.4385,
+      "step": 9238
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.154997322093474,
+      "learning_rate": 3.4238622609465936e-05,
+      "loss": 1.5625,
+      "step": 9239
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.742510022298713,
+      "learning_rate": 3.423474715125754e-05,
+      "loss": 1.2812,
+      "step": 9240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0302425527127195,
+      "learning_rate": 3.423087143604988e-05,
+      "loss": 0.7134,
+      "step": 9241
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4029381489629937,
+      "learning_rate": 3.422699546395084e-05,
+      "loss": 1.1699,
+      "step": 9242
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.178782791286578,
+      "learning_rate": 3.422311923506826e-05,
+      "loss": 1.1035,
+      "step": 9243
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.245130329625102,
+      "learning_rate": 3.4219242749510027e-05,
+      "loss": 1.0244,
+      "step": 9244
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7398858049879795,
+      "learning_rate": 3.4215366007384025e-05,
+      "loss": 0.6987,
+      "step": 9245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.544573873457425,
+      "learning_rate": 3.421148900879815e-05,
+      "loss": 0.9844,
+      "step": 9246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.755201800901081,
+      "learning_rate": 3.420761175386027e-05,
+      "loss": 0.8682,
+      "step": 9247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.970094252137366,
+      "learning_rate": 3.4203734242678306e-05,
+      "loss": 1.0205,
+      "step": 9248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.0512208246805175,
+      "learning_rate": 3.4199856475360166e-05,
+      "loss": 1.3281,
+      "step": 9249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.148504082775076,
+      "learning_rate": 3.419597845201377e-05,
+      "loss": 1.1191,
+      "step": 9250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.541377673739227,
+      "learning_rate": 3.419210017274703e-05,
+      "loss": 1.2285,
+      "step": 9251
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9456250359655916,
+      "learning_rate": 3.4188221637667893e-05,
+      "loss": 0.9062,
+      "step": 9252
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.692896370929076,
+      "learning_rate": 3.418434284688428e-05,
+      "loss": 1.6123,
+      "step": 9253
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0178526168174478,
+      "learning_rate": 3.418046380050415e-05,
+      "loss": 1.3135,
+      "step": 9254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.371546292050428,
+      "learning_rate": 3.4176584498635445e-05,
+      "loss": 1.249,
+      "step": 9255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2264262910640036,
+      "learning_rate": 3.417270494138613e-05,
+      "loss": 1.4766,
+      "step": 9256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.447054131561225,
+      "learning_rate": 3.4168825128864176e-05,
+      "loss": 1.3652,
+      "step": 9257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.0315326113731516,
+      "learning_rate": 3.4164945061177556e-05,
+      "loss": 1.208,
+      "step": 9258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.25535818692233,
+      "learning_rate": 3.416106473843423e-05,
+      "loss": 1.0645,
+      "step": 9259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.032922814086045,
+      "learning_rate": 3.415718416074221e-05,
+      "loss": 0.9609,
+      "step": 9260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.208715609685913,
+      "learning_rate": 3.415330332820949e-05,
+      "loss": 1.062,
+      "step": 9261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8459172636271655,
+      "learning_rate": 3.4149422240944054e-05,
+      "loss": 1.0684,
+      "step": 9262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.25824073247887,
+      "learning_rate": 3.4145540899053926e-05,
+      "loss": 1.1328,
+      "step": 9263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.42311285443457,
+      "learning_rate": 3.414165930264711e-05,
+      "loss": 0.3898,
+      "step": 9264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9550147090745873,
+      "learning_rate": 3.4137777451831654e-05,
+      "loss": 0.823,
+      "step": 9265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.137048654151727,
+      "learning_rate": 3.4133895346715564e-05,
+      "loss": 1.1465,
+      "step": 9266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9842620348895808,
+      "learning_rate": 3.413001298740688e-05,
+      "loss": 0.7915,
+      "step": 9267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8261759243232913,
+      "learning_rate": 3.4126130374013646e-05,
+      "loss": 1.1543,
+      "step": 9268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.757366491500129,
+      "learning_rate": 3.412224750664393e-05,
+      "loss": 1.1313,
+      "step": 9269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3941299575167183,
+      "learning_rate": 3.4118364385405775e-05,
+      "loss": 1.4717,
+      "step": 9270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.850935357513342,
+      "learning_rate": 3.4114481010407265e-05,
+      "loss": 1.0332,
+      "step": 9271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7707818009787877,
+      "learning_rate": 3.411059738175644e-05,
+      "loss": 1.2305,
+      "step": 9272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4104113575166104,
+      "learning_rate": 3.41067134995614e-05,
+      "loss": 1.1729,
+      "step": 9273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.366534298547935,
+      "learning_rate": 3.410282936393024e-05,
+      "loss": 0.5762,
+      "step": 9274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.176342083959238,
+      "learning_rate": 3.409894497497104e-05,
+      "loss": 1.5566,
+      "step": 9275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.180252003192207,
+      "learning_rate": 3.409506033279191e-05,
+      "loss": 1.5547,
+      "step": 9276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3742318868767773,
+      "learning_rate": 3.409117543750094e-05,
+      "loss": 0.8594,
+      "step": 9277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9128318826738337,
+      "learning_rate": 3.4087290289206275e-05,
+      "loss": 1.959,
+      "step": 9278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.231479399043929,
+      "learning_rate": 3.408340488801601e-05,
+      "loss": 1.8027,
+      "step": 9279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.146980707267569,
+      "learning_rate": 3.4079519234038297e-05,
+      "loss": 1.1152,
+      "step": 9280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.561314471407533,
+      "learning_rate": 3.407563332738125e-05,
+      "loss": 1.4219,
+      "step": 9281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2292053599579456,
+      "learning_rate": 3.4071747168153024e-05,
+      "loss": 0.665,
+      "step": 9282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.837360933747574,
+      "learning_rate": 3.4067860756461764e-05,
+      "loss": 1.5801,
+      "step": 9283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6801741355575426,
+      "learning_rate": 3.406397409241564e-05,
+      "loss": 1.083,
+      "step": 9284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.410689631311055,
+      "learning_rate": 3.40600871761228e-05,
+      "loss": 0.9473,
+      "step": 9285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7200542259420915,
+      "learning_rate": 3.4056200007691426e-05,
+      "loss": 0.5146,
+      "step": 9286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1953459496172756,
+      "learning_rate": 3.4052312587229686e-05,
+      "loss": 1.2949,
+      "step": 9287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8427305474316158,
+      "learning_rate": 3.404842491484578e-05,
+      "loss": 0.9399,
+      "step": 9288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6997571070438762,
+      "learning_rate": 3.4044536990647886e-05,
+      "loss": 1.1426,
+      "step": 9289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.150374999970588,
+      "learning_rate": 3.404064881474421e-05,
+      "loss": 1.377,
+      "step": 9290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.317569639935272,
+      "learning_rate": 3.4036760387242955e-05,
+      "loss": 1.1191,
+      "step": 9291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9337306069864884,
+      "learning_rate": 3.403287170825234e-05,
+      "loss": 1.1592,
+      "step": 9292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.746289090280741,
+      "learning_rate": 3.4028982777880594e-05,
+      "loss": 0.9443,
+      "step": 9293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2056964990447887,
+      "learning_rate": 3.4025093596235926e-05,
+      "loss": 1.0479,
+      "step": 9294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7647278618326787,
+      "learning_rate": 3.402120416342657e-05,
+      "loss": 1.064,
+      "step": 9295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.853194547352612,
+      "learning_rate": 3.401731447956078e-05,
+      "loss": 0.9629,
+      "step": 9296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1322669412753203,
+      "learning_rate": 3.40134245447468e-05,
+      "loss": 1.3457,
+      "step": 9297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.695033050645852,
+      "learning_rate": 3.400953435909288e-05,
+      "loss": 0.8364,
+      "step": 9298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.534697269218167,
+      "learning_rate": 3.40056439227073e-05,
+      "loss": 1.6738,
+      "step": 9299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.953712203162957,
+      "learning_rate": 3.400175323569829e-05,
+      "loss": 1.541,
+      "step": 9300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.978931013921931,
+      "learning_rate": 3.3997862298174176e-05,
+      "loss": 0.9468,
+      "step": 9301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.303439665666431,
+      "learning_rate": 3.39939711102432e-05,
+      "loss": 0.8672,
+      "step": 9302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.167767786802676,
+      "learning_rate": 3.399007967201369e-05,
+      "loss": 0.8301,
+      "step": 9303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5223943012576058,
+      "learning_rate": 3.398618798359391e-05,
+      "loss": 1.3867,
+      "step": 9304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.71847936225931,
+      "learning_rate": 3.398229604509217e-05,
+      "loss": 0.6318,
+      "step": 9305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5178164775467082,
+      "learning_rate": 3.3978403856616794e-05,
+      "loss": 0.374,
+      "step": 9306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3331428064034188,
+      "learning_rate": 3.39745114182761e-05,
+      "loss": 1.209,
+      "step": 9307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.134750771012766,
+      "learning_rate": 3.39706187301784e-05,
+      "loss": 0.6924,
+      "step": 9308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4096286361154147,
+      "learning_rate": 3.3966725792432034e-05,
+      "loss": 0.7988,
+      "step": 9309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3569419391263846,
+      "learning_rate": 3.396283260514534e-05,
+      "loss": 0.9058,
+      "step": 9310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.634109675575219,
+      "learning_rate": 3.3958939168426665e-05,
+      "loss": 0.4819,
+      "step": 9311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7552927904345095,
+      "learning_rate": 3.3955045482384355e-05,
+      "loss": 0.9587,
+      "step": 9312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.96319972938934,
+      "learning_rate": 3.395115154712678e-05,
+      "loss": 1.2168,
+      "step": 9313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.333969358770994,
+      "learning_rate": 3.39472573627623e-05,
+      "loss": 1.3389,
+      "step": 9314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.904890904822247,
+      "learning_rate": 3.394336292939929e-05,
+      "loss": 0.9785,
+      "step": 9315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5122944293158462,
+      "learning_rate": 3.393946824714614e-05,
+      "loss": 1.4287,
+      "step": 9316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9667142606353496,
+      "learning_rate": 3.3935573316111205e-05,
+      "loss": 0.7043,
+      "step": 9317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.271249365188021,
+      "learning_rate": 3.3931678136402925e-05,
+      "loss": 0.8075,
+      "step": 9318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.422285180581187,
+      "learning_rate": 3.392778270812966e-05,
+      "loss": 1.1113,
+      "step": 9319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7192002723792474,
+      "learning_rate": 3.392388703139985e-05,
+      "loss": 0.5732,
+      "step": 9320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.398965150079454,
+      "learning_rate": 3.391999110632189e-05,
+      "loss": 1.1328,
+      "step": 9321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.100777404940524,
+      "learning_rate": 3.391609493300422e-05,
+      "loss": 0.9219,
+      "step": 9322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.042071659525701,
+      "learning_rate": 3.391219851155524e-05,
+      "loss": 1.0518,
+      "step": 9323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2415971676666313,
+      "learning_rate": 3.390830184208341e-05,
+      "loss": 1.4199,
+      "step": 9324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1351680889424176,
+      "learning_rate": 3.390440492469717e-05,
+      "loss": 0.6976,
+      "step": 9325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.099815708763651,
+      "learning_rate": 3.390050775950497e-05,
+      "loss": 1.2129,
+      "step": 9326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6950631288909124,
+      "learning_rate": 3.389661034661525e-05,
+      "loss": 1.126,
+      "step": 9327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.513390917367097,
+      "learning_rate": 3.38927126861365e-05,
+      "loss": 1.1416,
+      "step": 9328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3424888778308492,
+      "learning_rate": 3.388881477817717e-05,
+      "loss": 1.3262,
+      "step": 9329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.003954363277995,
+      "learning_rate": 3.3884916622845744e-05,
+      "loss": 0.8652,
+      "step": 9330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7227126434484314,
+      "learning_rate": 3.3881018220250696e-05,
+      "loss": 1.0532,
+      "step": 9331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.515663248489853,
+      "learning_rate": 3.387711957050054e-05,
+      "loss": 0.9834,
+      "step": 9332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.250899288329784,
+      "learning_rate": 3.3873220673703756e-05,
+      "loss": 1.3008,
+      "step": 9333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.9102628782798865,
+      "learning_rate": 3.386932152996885e-05,
+      "loss": 1.4346,
+      "step": 9334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.8684975130779655,
+      "learning_rate": 3.386542213940434e-05,
+      "loss": 0.9746,
+      "step": 9335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.059330321800019,
+      "learning_rate": 3.386152250211874e-05,
+      "loss": 1.1191,
+      "step": 9336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.193745297288603,
+      "learning_rate": 3.385762261822058e-05,
+      "loss": 0.9883,
+      "step": 9337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.027914247791902,
+      "learning_rate": 3.3853722487818374e-05,
+      "loss": 1.5391,
+      "step": 9338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.69182212506637,
+      "learning_rate": 3.384982211102069e-05,
+      "loss": 1.2852,
+      "step": 9339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.430790246597818,
+      "learning_rate": 3.384592148793606e-05,
+      "loss": 1.3184,
+      "step": 9340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.16582163193453,
+      "learning_rate": 3.3842020618673034e-05,
+      "loss": 1.4355,
+      "step": 9341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8375559713960568,
+      "learning_rate": 3.383811950334016e-05,
+      "loss": 0.5879,
+      "step": 9342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5508125618711306,
+      "learning_rate": 3.3834218142046034e-05,
+      "loss": 1.3467,
+      "step": 9343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.606159170200923,
+      "learning_rate": 3.383031653489921e-05,
+      "loss": 0.7686,
+      "step": 9344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.34550795165721,
+      "learning_rate": 3.3826414682008266e-05,
+      "loss": 0.9785,
+      "step": 9345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6883547111302937,
+      "learning_rate": 3.38225125834818e-05,
+      "loss": 1.7148,
+      "step": 9346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.354580887200661,
+      "learning_rate": 3.38186102394284e-05,
+      "loss": 1.3838,
+      "step": 9347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5775832098212166,
+      "learning_rate": 3.381470764995667e-05,
+      "loss": 0.7852,
+      "step": 9348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.768171646051829,
+      "learning_rate": 3.3810804815175214e-05,
+      "loss": 1.6484,
+      "step": 9349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.086384128946659,
+      "learning_rate": 3.380690173519264e-05,
+      "loss": 0.5479,
+      "step": 9350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.168434810590373,
+      "learning_rate": 3.3802998410117574e-05,
+      "loss": 1.5469,
+      "step": 9351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2430207834084475,
+      "learning_rate": 3.379909484005865e-05,
+      "loss": 1.3086,
+      "step": 9352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7201001552767,
+      "learning_rate": 3.379519102512451e-05,
+      "loss": 0.7148,
+      "step": 9353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.667668908441361,
+      "learning_rate": 3.379128696542378e-05,
+      "loss": 1.6211,
+      "step": 9354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6264441671792573,
+      "learning_rate": 3.37873826610651e-05,
+      "loss": 0.8682,
+      "step": 9355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.432314629248577,
+      "learning_rate": 3.378347811215715e-05,
+      "loss": 1.582,
+      "step": 9356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9663838290913012,
+      "learning_rate": 3.377957331880857e-05,
+      "loss": 1.4023,
+      "step": 9357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4031461520627944,
+      "learning_rate": 3.377566828112805e-05,
+      "loss": 1.1504,
+      "step": 9358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8078777872717833,
+      "learning_rate": 3.3771762999224246e-05,
+      "loss": 0.7778,
+      "step": 9359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6151016349808085,
+      "learning_rate": 3.3767857473205854e-05,
+      "loss": 1.1719,
+      "step": 9360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2172418761087864,
+      "learning_rate": 3.376395170318155e-05,
+      "loss": 1.1094,
+      "step": 9361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1081498720865914,
+      "learning_rate": 3.376004568926004e-05,
+      "loss": 1.4629,
+      "step": 9362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.668403172424002,
+      "learning_rate": 3.3756139431550026e-05,
+      "loss": 1.1345,
+      "step": 9363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3917414501546643,
+      "learning_rate": 3.3752232930160225e-05,
+      "loss": 1.25,
+      "step": 9364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.692037086474131,
+      "learning_rate": 3.374832618519934e-05,
+      "loss": 1.1162,
+      "step": 9365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0180061703385666,
+      "learning_rate": 3.374441919677609e-05,
+      "loss": 1.1846,
+      "step": 9366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.151627492373222,
+      "learning_rate": 3.374051196499922e-05,
+      "loss": 0.8848,
+      "step": 9367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.344462921593533,
+      "learning_rate": 3.3736604489977466e-05,
+      "loss": 1.3477,
+      "step": 9368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5020860551813477,
+      "learning_rate": 3.373269677181956e-05,
+      "loss": 1.377,
+      "step": 9369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7326311054520125,
+      "learning_rate": 3.372878881063425e-05,
+      "loss": 1.0312,
+      "step": 9370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.238756340233555,
+      "learning_rate": 3.3724880606530314e-05,
+      "loss": 1.334,
+      "step": 9371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.291950994250616,
+      "learning_rate": 3.3720972159616496e-05,
+      "loss": 0.835,
+      "step": 9372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.051004049368628,
+      "learning_rate": 3.3717063470001576e-05,
+      "loss": 1.3232,
+      "step": 9373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.1900407934331305,
+      "learning_rate": 3.3713154537794325e-05,
+      "loss": 1.1294,
+      "step": 9374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8545862339643326,
+      "learning_rate": 3.370924536310353e-05,
+      "loss": 0.7656,
+      "step": 9375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7114252352177766,
+      "learning_rate": 3.3705335946037986e-05,
+      "loss": 0.3525,
+      "step": 9376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1427673320818825,
+      "learning_rate": 3.3701426286706484e-05,
+      "loss": 1.1191,
+      "step": 9377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0793795133311828,
+      "learning_rate": 3.369751638521783e-05,
+      "loss": 1.1777,
+      "step": 9378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9923654654712784,
+      "learning_rate": 3.3693606241680836e-05,
+      "loss": 0.4482,
+      "step": 9379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.090638354041419,
+      "learning_rate": 3.368969585620432e-05,
+      "loss": 0.7222,
+      "step": 9380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.280517641717482,
+      "learning_rate": 3.368578522889711e-05,
+      "loss": 0.5645,
+      "step": 9381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3506744559860717,
+      "learning_rate": 3.368187435986803e-05,
+      "loss": 1.4766,
+      "step": 9382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3637608065224494,
+      "learning_rate": 3.3677963249225916e-05,
+      "loss": 1.1895,
+      "step": 9383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.0481775021442905,
+      "learning_rate": 3.367405189707963e-05,
+      "loss": 1.127,
+      "step": 9384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.016227711791414,
+      "learning_rate": 3.3670140303538e-05,
+      "loss": 1.5322,
+      "step": 9385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.345456188933071,
+      "learning_rate": 3.36662284687099e-05,
+      "loss": 1.7695,
+      "step": 9386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0140707329320717,
+      "learning_rate": 3.366231639270419e-05,
+      "loss": 1.2383,
+      "step": 9387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.518306264178217,
+      "learning_rate": 3.365840407562974e-05,
+      "loss": 1.5137,
+      "step": 9388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.751206394693363,
+      "learning_rate": 3.365449151759542e-05,
+      "loss": 1.4922,
+      "step": 9389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.933453980605755,
+      "learning_rate": 3.365057871871014e-05,
+      "loss": 0.9316,
+      "step": 9390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.435911942126532,
+      "learning_rate": 3.364666567908276e-05,
+      "loss": 1.1719,
+      "step": 9391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7444949114297956,
+      "learning_rate": 3.364275239882221e-05,
+      "loss": 1.252,
+      "step": 9392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0395518917768403,
+      "learning_rate": 3.3638838878037366e-05,
+      "loss": 0.8975,
+      "step": 9393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.020404216965836,
+      "learning_rate": 3.3634925116837164e-05,
+      "loss": 2.2246,
+      "step": 9394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.960897772775995,
+      "learning_rate": 3.36310111153305e-05,
+      "loss": 0.9346,
+      "step": 9395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4922186870477128,
+      "learning_rate": 3.3627096873626324e-05,
+      "loss": 0.9028,
+      "step": 9396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8578169299286027,
+      "learning_rate": 3.362318239183354e-05,
+      "loss": 1.335,
+      "step": 9397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2439459839981803,
+      "learning_rate": 3.361926767006111e-05,
+      "loss": 0.7285,
+      "step": 9398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.074720576716925,
+      "learning_rate": 3.361535270841797e-05,
+      "loss": 1.1191,
+      "step": 9399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.009419506814367,
+      "learning_rate": 3.361143750701307e-05,
+      "loss": 1.2715,
+      "step": 9400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2180314234117464,
+      "learning_rate": 3.3607522065955365e-05,
+      "loss": 0.8281,
+      "step": 9401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6480677415263463,
+      "learning_rate": 3.3603606385353825e-05,
+      "loss": 1.209,
+      "step": 9402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.525442549193092,
+      "learning_rate": 3.359969046531743e-05,
+      "loss": 1.2598,
+      "step": 9403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.866437351310727,
+      "learning_rate": 3.359577430595514e-05,
+      "loss": 1.002,
+      "step": 9404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.624978298569249,
+      "learning_rate": 3.359185790737596e-05,
+      "loss": 0.9521,
+      "step": 9405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.592173354895486,
+      "learning_rate": 3.358794126968887e-05,
+      "loss": 0.8994,
+      "step": 9406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.694296913652472,
+      "learning_rate": 3.3584024393002865e-05,
+      "loss": 1.1367,
+      "step": 9407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6460027164897975,
+      "learning_rate": 3.358010727742696e-05,
+      "loss": 0.8955,
+      "step": 9408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.343418245888349,
+      "learning_rate": 3.357618992307016e-05,
+      "loss": 0.9883,
+      "step": 9409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.901912327569298,
+      "learning_rate": 3.357227233004149e-05,
+      "loss": 1.3193,
+      "step": 9410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.629857064998138,
+      "learning_rate": 3.356835449844997e-05,
+      "loss": 0.4578,
+      "step": 9411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2596941963848862,
+      "learning_rate": 3.3564436428404625e-05,
+      "loss": 1.0049,
+      "step": 9412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.90059563011814,
+      "learning_rate": 3.3560518120014514e-05,
+      "loss": 0.9497,
+      "step": 9413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6061800733427765,
+      "learning_rate": 3.355659957338866e-05,
+      "loss": 1.0098,
+      "step": 9414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8426825228076513,
+      "learning_rate": 3.3552680788636124e-05,
+      "loss": 0.8652,
+      "step": 9415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.401547977798226,
+      "learning_rate": 3.354876176586596e-05,
+      "loss": 1.0381,
+      "step": 9416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.173892577470566,
+      "learning_rate": 3.354484250518724e-05,
+      "loss": 0.957,
+      "step": 9417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.879433401077605,
+      "learning_rate": 3.3540923006709035e-05,
+      "loss": 1.3984,
+      "step": 9418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8473467874544793,
+      "learning_rate": 3.3537003270540415e-05,
+      "loss": 1.1992,
+      "step": 9419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7061977140918176,
+      "learning_rate": 3.353308329679047e-05,
+      "loss": 0.6598,
+      "step": 9420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0914583291927538,
+      "learning_rate": 3.3529163085568294e-05,
+      "loss": 1.373,
+      "step": 9421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6419681541493008,
+      "learning_rate": 3.3525242636982976e-05,
+      "loss": 0.7168,
+      "step": 9422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.528108532903472,
+      "learning_rate": 3.352132195114363e-05,
+      "loss": 0.8955,
+      "step": 9423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.058147217564399,
+      "learning_rate": 3.351740102815937e-05,
+      "loss": 1.3301,
+      "step": 9424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.09737323259995,
+      "learning_rate": 3.35134798681393e-05,
+      "loss": 0.498,
+      "step": 9425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4233185434567868,
+      "learning_rate": 3.350955847119254e-05,
+      "loss": 1.1992,
+      "step": 9426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7202715069236407,
+      "learning_rate": 3.350563683742824e-05,
+      "loss": 0.6433,
+      "step": 9427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.195896896492002,
+      "learning_rate": 3.350171496695554e-05,
+      "loss": 1.6465,
+      "step": 9428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.628192348398849,
+      "learning_rate": 3.349779285988356e-05,
+      "loss": 1.8066,
+      "step": 9429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5113269743603,
+      "learning_rate": 3.3493870516321475e-05,
+      "loss": 0.9531,
+      "step": 9430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.849536220753123,
+      "learning_rate": 3.348994793637842e-05,
+      "loss": 1.5391,
+      "step": 9431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.707643283926289,
+      "learning_rate": 3.348602512016358e-05,
+      "loss": 0.9756,
+      "step": 9432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.801293404756537,
+      "learning_rate": 3.348210206778612e-05,
+      "loss": 1.2529,
+      "step": 9433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3161081569367408,
+      "learning_rate": 3.34781787793552e-05,
+      "loss": 0.9404,
+      "step": 9434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.679496755006558,
+      "learning_rate": 3.347425525498002e-05,
+      "loss": 1.6836,
+      "step": 9435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.456733819034262,
+      "learning_rate": 3.347033149476976e-05,
+      "loss": 1.4121,
+      "step": 9436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7865425073381953,
+      "learning_rate": 3.3466407498833635e-05,
+      "loss": 1.0234,
+      "step": 9437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8735543638509498,
+      "learning_rate": 3.3462483267280834e-05,
+      "loss": 0.6812,
+      "step": 9438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8838026212431,
+      "learning_rate": 3.345855880022056e-05,
+      "loss": 1.0918,
+      "step": 9439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.966813343338846,
+      "learning_rate": 3.345463409776204e-05,
+      "loss": 0.8789,
+      "step": 9440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.990537514819139,
+      "learning_rate": 3.3450709160014505e-05,
+      "loss": 1.583,
+      "step": 9441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5535651958591488,
+      "learning_rate": 3.344678398708716e-05,
+      "loss": 1.541,
+      "step": 9442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.414723574667236,
+      "learning_rate": 3.344285857908927e-05,
+      "loss": 1.0195,
+      "step": 9443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4497011430596216,
+      "learning_rate": 3.343893293613005e-05,
+      "loss": 1.1504,
+      "step": 9444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6750200583552228,
+      "learning_rate": 3.343500705831877e-05,
+      "loss": 1.1069,
+      "step": 9445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6810120625948426,
+      "learning_rate": 3.343108094576467e-05,
+      "loss": 1.0889,
+      "step": 9446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8711412490756274,
+      "learning_rate": 3.3427154598577027e-05,
+      "loss": 1.4043,
+      "step": 9447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2084221930810752,
+      "learning_rate": 3.34232280168651e-05,
+      "loss": 1.2275,
+      "step": 9448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.242265956989154,
+      "learning_rate": 3.3419301200738165e-05,
+      "loss": 0.9199,
+      "step": 9449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.548111082623918,
+      "learning_rate": 3.34153741503055e-05,
+      "loss": 0.9141,
+      "step": 9450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8111533437814127,
+      "learning_rate": 3.341144686567641e-05,
+      "loss": 0.5469,
+      "step": 9451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.096389359520365,
+      "learning_rate": 3.340751934696017e-05,
+      "loss": 0.7915,
+      "step": 9452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2998913891576516,
+      "learning_rate": 3.340359159426609e-05,
+      "loss": 0.5974,
+      "step": 9453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0873500884358416,
+      "learning_rate": 3.339966360770349e-05,
+      "loss": 1.0244,
+      "step": 9454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.769179208182011,
+      "learning_rate": 3.3395735387381654e-05,
+      "loss": 1.125,
+      "step": 9455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.648535390190047,
+      "learning_rate": 3.339180693340993e-05,
+      "loss": 0.957,
+      "step": 9456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.704818371582146,
+      "learning_rate": 3.338787824589763e-05,
+      "loss": 1.6914,
+      "step": 9457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.850255652341846,
+      "learning_rate": 3.338394932495409e-05,
+      "loss": 1.1472,
+      "step": 9458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.661172821603705,
+      "learning_rate": 3.3380020170688655e-05,
+      "loss": 0.5342,
+      "step": 9459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.04262351625614,
+      "learning_rate": 3.337609078321068e-05,
+      "loss": 1.1973,
+      "step": 9460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.831726647142726,
+      "learning_rate": 3.33721611626295e-05,
+      "loss": 0.9932,
+      "step": 9461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5105661265939556,
+      "learning_rate": 3.336823130905449e-05,
+      "loss": 1.0215,
+      "step": 9462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.439484595566913,
+      "learning_rate": 3.3364301222595e-05,
+      "loss": 1.1455,
+      "step": 9463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.5188670369544175,
+      "learning_rate": 3.336037090336042e-05,
+      "loss": 1.0518,
+      "step": 9464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.23109814229042,
+      "learning_rate": 3.3356440351460113e-05,
+      "loss": 1.4814,
+      "step": 9465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.258072983359516,
+      "learning_rate": 3.3352509567003476e-05,
+      "loss": 0.6953,
+      "step": 9466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.197334618811125,
+      "learning_rate": 3.33485785500999e-05,
+      "loss": 0.9316,
+      "step": 9467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.028294463465909,
+      "learning_rate": 3.334464730085878e-05,
+      "loss": 1.459,
+      "step": 9468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1264838701128275,
+      "learning_rate": 3.3340715819389524e-05,
+      "loss": 1.1416,
+      "step": 9469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.588739358170029,
+      "learning_rate": 3.333678410580154e-05,
+      "loss": 1.0225,
+      "step": 9470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.209958564059154,
+      "learning_rate": 3.333285216020425e-05,
+      "loss": 0.9268,
+      "step": 9471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7046579013061987,
+      "learning_rate": 3.332891998270708e-05,
+      "loss": 0.7275,
+      "step": 9472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5937946729374874,
+      "learning_rate": 3.332498757341945e-05,
+      "loss": 0.4526,
+      "step": 9473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.044004397493967,
+      "learning_rate": 3.3321054932450815e-05,
+      "loss": 1.1289,
+      "step": 9474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4959763150231207,
+      "learning_rate": 3.3317122059910607e-05,
+      "loss": 1.0068,
+      "step": 9475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.992082301335512,
+      "learning_rate": 3.331318895590827e-05,
+      "loss": 1.3223,
+      "step": 9476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3272459687650766,
+      "learning_rate": 3.330925562055327e-05,
+      "loss": 1.5273,
+      "step": 9477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.564526992259361,
+      "learning_rate": 3.3305322053955065e-05,
+      "loss": 0.9546,
+      "step": 9478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.820511744850238,
+      "learning_rate": 3.330138825622314e-05,
+      "loss": 0.834,
+      "step": 9479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.895179667367812,
+      "learning_rate": 3.3297454227466956e-05,
+      "loss": 1.3359,
+      "step": 9480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.620183886270174,
+      "learning_rate": 3.3293519967795994e-05,
+      "loss": 0.9736,
+      "step": 9481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.0858077499407,
+      "learning_rate": 3.328958547731975e-05,
+      "loss": 0.7217,
+      "step": 9482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.093340250032179,
+      "learning_rate": 3.3285650756147724e-05,
+      "loss": 0.6912,
+      "step": 9483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0821079877962583,
+      "learning_rate": 3.3281715804389403e-05,
+      "loss": 0.5061,
+      "step": 9484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.811714486524058,
+      "learning_rate": 3.3277780622154306e-05,
+      "loss": 1.0205,
+      "step": 9485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.196399134705609,
+      "learning_rate": 3.327384520955194e-05,
+      "loss": 1.0996,
+      "step": 9486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.652895346183855,
+      "learning_rate": 3.3269909566691824e-05,
+      "loss": 1.584,
+      "step": 9487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1085730050337426,
+      "learning_rate": 3.32659736936835e-05,
+      "loss": 0.7437,
+      "step": 9488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.7688489671189425,
+      "learning_rate": 3.3262037590636497e-05,
+      "loss": 1.375,
+      "step": 9489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0517954685176787,
+      "learning_rate": 3.3258101257660343e-05,
+      "loss": 0.749,
+      "step": 9490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.144366722184959,
+      "learning_rate": 3.3254164694864595e-05,
+      "loss": 0.8975,
+      "step": 9491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.093448508628697,
+      "learning_rate": 3.325022790235881e-05,
+      "loss": 1.1006,
+      "step": 9492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.060728728701325,
+      "learning_rate": 3.324629088025254e-05,
+      "loss": 1.3345,
+      "step": 9493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0718319481336738,
+      "learning_rate": 3.3242353628655355e-05,
+      "loss": 0.7612,
+      "step": 9494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9775829470439323,
+      "learning_rate": 3.323841614767681e-05,
+      "loss": 1.2283,
+      "step": 9495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.764896179834722,
+      "learning_rate": 3.323447843742651e-05,
+      "loss": 0.8818,
+      "step": 9496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5248599929518445,
+      "learning_rate": 3.323054049801402e-05,
+      "loss": 1.3555,
+      "step": 9497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.151465595088182,
+      "learning_rate": 3.322660232954894e-05,
+      "loss": 0.8115,
+      "step": 9498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.408330331571792,
+      "learning_rate": 3.322266393214086e-05,
+      "loss": 1.0938,
+      "step": 9499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3334318775984393,
+      "learning_rate": 3.32187253058994e-05,
+      "loss": 1.167,
+      "step": 9500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.437685363993955,
+      "learning_rate": 3.321478645093415e-05,
+      "loss": 1.6816,
+      "step": 9501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.676105865416021,
+      "learning_rate": 3.321084736735476e-05,
+      "loss": 2.0312,
+      "step": 9502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2285692240281434,
+      "learning_rate": 3.3206908055270815e-05,
+      "loss": 1.0791,
+      "step": 9503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.938809366798868,
+      "learning_rate": 3.320296851479196e-05,
+      "loss": 1.1885,
+      "step": 9504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.797071223019895,
+      "learning_rate": 3.319902874602783e-05,
+      "loss": 1.0469,
+      "step": 9505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.185107203108055,
+      "learning_rate": 3.3195088749088065e-05,
+      "loss": 1.1582,
+      "step": 9506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.95829212100342,
+      "learning_rate": 3.319114852408233e-05,
+      "loss": 0.9927,
+      "step": 9507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.00582152171596,
+      "learning_rate": 3.318720807112025e-05,
+      "loss": 0.9277,
+      "step": 9508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7115640493372912,
+      "learning_rate": 3.318326739031151e-05,
+      "loss": 0.8867,
+      "step": 9509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.988244676574223,
+      "learning_rate": 3.3179326481765764e-05,
+      "loss": 2.082,
+      "step": 9510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7109101483726916,
+      "learning_rate": 3.31753853455927e-05,
+      "loss": 0.9331,
+      "step": 9511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.562196015890681,
+      "learning_rate": 3.317144398190198e-05,
+      "loss": 0.7876,
+      "step": 9512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.246443385011842,
+      "learning_rate": 3.316750239080332e-05,
+      "loss": 1.1055,
+      "step": 9513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5047247195467786,
+      "learning_rate": 3.316356057240637e-05,
+      "loss": 0.3418,
+      "step": 9514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.526414561850465,
+      "learning_rate": 3.3159618526820865e-05,
+      "loss": 0.7158,
+      "step": 9515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.73187653236833,
+      "learning_rate": 3.31556762541565e-05,
+      "loss": 1.4844,
+      "step": 9516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.706695924583323,
+      "learning_rate": 3.3151733754522984e-05,
+      "loss": 0.7754,
+      "step": 9517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.347562308820093,
+      "learning_rate": 3.314779102803003e-05,
+      "loss": 1.5801,
+      "step": 9518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.524009655227949,
+      "learning_rate": 3.3143848074787375e-05,
+      "loss": 0.8281,
+      "step": 9519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.372201575914758,
+      "learning_rate": 3.313990489490474e-05,
+      "loss": 1.1406,
+      "step": 9520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.010791293082312,
+      "learning_rate": 3.3135961488491865e-05,
+      "loss": 1.2578,
+      "step": 9521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.664664953442926,
+      "learning_rate": 3.3132017855658494e-05,
+      "loss": 1.3359,
+      "step": 9522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.070661141594198,
+      "learning_rate": 3.312807399651437e-05,
+      "loss": 0.9629,
+      "step": 9523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3630480942902867,
+      "learning_rate": 3.3124129911169275e-05,
+      "loss": 1.04,
+      "step": 9524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.969633237323409,
+      "learning_rate": 3.312018559973294e-05,
+      "loss": 0.9189,
+      "step": 9525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4696689295803638,
+      "learning_rate": 3.311624106231515e-05,
+      "loss": 1.3457,
+      "step": 9526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0729317012785455,
+      "learning_rate": 3.311229629902567e-05,
+      "loss": 1.0117,
+      "step": 9527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.878067039212629,
+      "learning_rate": 3.3108351309974284e-05,
+      "loss": 1.293,
+      "step": 9528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.985973149939672,
+      "learning_rate": 3.310440609527079e-05,
+      "loss": 1.1367,
+      "step": 9529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.494895346061085,
+      "learning_rate": 3.310046065502497e-05,
+      "loss": 1.2168,
+      "step": 9530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.818946315090767,
+      "learning_rate": 3.309651498934662e-05,
+      "loss": 1.1348,
+      "step": 9531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6568848148488717,
+      "learning_rate": 3.309256909834556e-05,
+      "loss": 1.1094,
+      "step": 9532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5533395076743393,
+      "learning_rate": 3.30886229821316e-05,
+      "loss": 1.3828,
+      "step": 9533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.053420041229308,
+      "learning_rate": 3.3084676640814556e-05,
+      "loss": 1.3851,
+      "step": 9534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.4273376278637855,
+      "learning_rate": 3.3080730074504244e-05,
+      "loss": 0.8794,
+      "step": 9535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.958188398953659,
+      "learning_rate": 3.307678328331051e-05,
+      "loss": 0.8955,
+      "step": 9536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.517437317101415,
+      "learning_rate": 3.307283626734318e-05,
+      "loss": 1.25,
+      "step": 9537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2207635303613737,
+      "learning_rate": 3.3068889026712113e-05,
+      "loss": 1.1855,
+      "step": 9538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1001743425932786,
+      "learning_rate": 3.306494156152714e-05,
+      "loss": 1.2617,
+      "step": 9539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.134724971736653,
+      "learning_rate": 3.3060993871898136e-05,
+      "loss": 1.0781,
+      "step": 9540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.903291734849263,
+      "learning_rate": 3.3057045957934946e-05,
+      "loss": 0.5012,
+      "step": 9541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5707083384575222,
+      "learning_rate": 3.305309781974745e-05,
+      "loss": 0.8511,
+      "step": 9542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2342853661544915,
+      "learning_rate": 3.304914945744552e-05,
+      "loss": 1.5918,
+      "step": 9543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.06724248050982,
+      "learning_rate": 3.304520087113904e-05,
+      "loss": 1.377,
+      "step": 9544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.09969226762242,
+      "learning_rate": 3.3041252060937895e-05,
+      "loss": 0.8564,
+      "step": 9545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.473595613009767,
+      "learning_rate": 3.303730302695196e-05,
+      "loss": 0.6475,
+      "step": 9546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6400326340615126,
+      "learning_rate": 3.303335376929118e-05,
+      "loss": 0.5714,
+      "step": 9547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.940199395957929,
+      "learning_rate": 3.3029404288065426e-05,
+      "loss": 1.0781,
+      "step": 9548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9439077088049865,
+      "learning_rate": 3.3025454583384617e-05,
+      "loss": 1.2754,
+      "step": 9549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9321596776103345,
+      "learning_rate": 3.302150465535867e-05,
+      "loss": 1.2861,
+      "step": 9550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.184947360310442,
+      "learning_rate": 3.301755450409752e-05,
+      "loss": 1.332,
+      "step": 9551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.742756468941746,
+      "learning_rate": 3.301360412971109e-05,
+      "loss": 1.0527,
+      "step": 9552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.523333138166258,
+      "learning_rate": 3.3009653532309326e-05,
+      "loss": 1.2168,
+      "step": 9553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0677024094287475,
+      "learning_rate": 3.3005702712002153e-05,
+      "loss": 0.9951,
+      "step": 9554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.1939010644026835,
+      "learning_rate": 3.300175166889954e-05,
+      "loss": 1.498,
+      "step": 9555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1155316961948314,
+      "learning_rate": 3.2997800403111434e-05,
+      "loss": 0.5449,
+      "step": 9556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.005589394699085,
+      "learning_rate": 3.2993848914747804e-05,
+      "loss": 0.6904,
+      "step": 9557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.33504948500886,
+      "learning_rate": 3.298989720391861e-05,
+      "loss": 0.6821,
+      "step": 9558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.08932569256537,
+      "learning_rate": 3.298594527073382e-05,
+      "loss": 1.2773,
+      "step": 9559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.843157192944404,
+      "learning_rate": 3.2981993115303435e-05,
+      "loss": 1.2402,
+      "step": 9560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.219910546468048,
+      "learning_rate": 3.2978040737737424e-05,
+      "loss": 1.5586,
+      "step": 9561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.705294934339451,
+      "learning_rate": 3.2974088138145795e-05,
+      "loss": 1.7324,
+      "step": 9562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.003036776783258,
+      "learning_rate": 3.297013531663854e-05,
+      "loss": 0.8691,
+      "step": 9563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.991759070909045,
+      "learning_rate": 3.296618227332566e-05,
+      "loss": 1.249,
+      "step": 9564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.955647785790132,
+      "learning_rate": 3.2962229008317166e-05,
+      "loss": 1.5117,
+      "step": 9565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.041631794211393,
+      "learning_rate": 3.2958275521723084e-05,
+      "loss": 1.1055,
+      "step": 9566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.646921369687667,
+      "learning_rate": 3.295432181365343e-05,
+      "loss": 0.793,
+      "step": 9567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6144199598925515,
+      "learning_rate": 3.295036788421825e-05,
+      "loss": 0.542,
+      "step": 9568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7823214235265303,
+      "learning_rate": 3.2946413733527546e-05,
+      "loss": 1.252,
+      "step": 9569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.997579551831054,
+      "learning_rate": 3.29424593616914e-05,
+      "loss": 0.8066,
+      "step": 9570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.855412397056215,
+      "learning_rate": 3.293850476881984e-05,
+      "loss": 1.8223,
+      "step": 9571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.016080956604336,
+      "learning_rate": 3.293454995502293e-05,
+      "loss": 1.2402,
+      "step": 9572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.473040847638416,
+      "learning_rate": 3.2930594920410717e-05,
+      "loss": 0.7168,
+      "step": 9573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4882294126946927,
+      "learning_rate": 3.2926639665093276e-05,
+      "loss": 0.6689,
+      "step": 9574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.677497536017464,
+      "learning_rate": 3.292268418918069e-05,
+      "loss": 0.6357,
+      "step": 9575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.267040152727152,
+      "learning_rate": 3.291872849278302e-05,
+      "loss": 1.1602,
+      "step": 9576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.702313592749753,
+      "learning_rate": 3.291477257601036e-05,
+      "loss": 1.5464,
+      "step": 9577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.597118705926071,
+      "learning_rate": 3.29108164389728e-05,
+      "loss": 0.7812,
+      "step": 9578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.271882144544195,
+      "learning_rate": 3.290686008178045e-05,
+      "loss": 0.8008,
+      "step": 9579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.375858352549022,
+      "learning_rate": 3.290290350454338e-05,
+      "loss": 0.8948,
+      "step": 9580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.748403172281317,
+      "learning_rate": 3.289894670737174e-05,
+      "loss": 1.1714,
+      "step": 9581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1092057901966728,
+      "learning_rate": 3.2894989690375626e-05,
+      "loss": 0.7271,
+      "step": 9582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7896726993284826,
+      "learning_rate": 3.289103245366516e-05,
+      "loss": 1.0439,
+      "step": 9583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.261378595499312,
+      "learning_rate": 3.288707499735048e-05,
+      "loss": 0.8833,
+      "step": 9584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.387266782121288,
+      "learning_rate": 3.2883117321541706e-05,
+      "loss": 1.1816,
+      "step": 9585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.479742615380048,
+      "learning_rate": 3.287915942634899e-05,
+      "loss": 1.1602,
+      "step": 9586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.150888695934921,
+      "learning_rate": 3.2875201311882474e-05,
+      "loss": 0.9639,
+      "step": 9587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.816129802131667,
+      "learning_rate": 3.28712429782523e-05,
+      "loss": 1.0098,
+      "step": 9588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1614442265674514,
+      "learning_rate": 3.2867284425568664e-05,
+      "loss": 1.4346,
+      "step": 9589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.252228301504261,
+      "learning_rate": 3.2863325653941686e-05,
+      "loss": 0.8037,
+      "step": 9590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6376038372697996,
+      "learning_rate": 3.2859366663481564e-05,
+      "loss": 0.8472,
+      "step": 9591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.238116086211547,
+      "learning_rate": 3.285540745429847e-05,
+      "loss": 0.9678,
+      "step": 9592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.15098448547273,
+      "learning_rate": 3.285144802650257e-05,
+      "loss": 1.3477,
+      "step": 9593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7640423728801558,
+      "learning_rate": 3.284748838020408e-05,
+      "loss": 1.4688,
+      "step": 9594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2602517438455036,
+      "learning_rate": 3.284352851551318e-05,
+      "loss": 1.4238,
+      "step": 9595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.542312538212154,
+      "learning_rate": 3.283956843254007e-05,
+      "loss": 1.459,
+      "step": 9596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.164353809580447,
+      "learning_rate": 3.2835608131394966e-05,
+      "loss": 0.874,
+      "step": 9597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7545147162747328,
+      "learning_rate": 3.283164761218808e-05,
+      "loss": 1.5059,
+      "step": 9598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5056358707503312,
+      "learning_rate": 3.282768687502963e-05,
+      "loss": 0.7144,
+      "step": 9599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.398712497853622,
+      "learning_rate": 3.282372592002984e-05,
+      "loss": 1.0449,
+      "step": 9600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.151623208517716,
+      "learning_rate": 3.281976474729894e-05,
+      "loss": 1.4238,
+      "step": 9601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.281576263925427,
+      "learning_rate": 3.2815803356947165e-05,
+      "loss": 0.6184,
+      "step": 9602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.924766226606325,
+      "learning_rate": 3.281184174908477e-05,
+      "loss": 0.9902,
+      "step": 9603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.343368605895693,
+      "learning_rate": 3.2807879923822004e-05,
+      "loss": 0.7725,
+      "step": 9604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8727489946592069,
+      "learning_rate": 3.280391788126912e-05,
+      "loss": 1.3926,
+      "step": 9605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.297632933996951,
+      "learning_rate": 3.279995562153636e-05,
+      "loss": 0.9219,
+      "step": 9606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8556150302391488,
+      "learning_rate": 3.279599314473403e-05,
+      "loss": 0.8457,
+      "step": 9607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1622896610344675,
+      "learning_rate": 3.279203045097238e-05,
+      "loss": 0.9678,
+      "step": 9608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3954635224133836,
+      "learning_rate": 3.278806754036169e-05,
+      "loss": 0.9639,
+      "step": 9609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.707999008279954,
+      "learning_rate": 3.278410441301225e-05,
+      "loss": 0.8779,
+      "step": 9610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.9790135912453355,
+      "learning_rate": 3.278014106903436e-05,
+      "loss": 1.0518,
+      "step": 9611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.416075876897123,
+      "learning_rate": 3.2776177508538304e-05,
+      "loss": 1.3447,
+      "step": 9612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1591409294138417,
+      "learning_rate": 3.27722137316344e-05,
+      "loss": 1.1094,
+      "step": 9613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.593065146828405,
+      "learning_rate": 3.276824973843295e-05,
+      "loss": 0.83,
+      "step": 9614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.289484499417126,
+      "learning_rate": 3.276428552904427e-05,
+      "loss": 1.166,
+      "step": 9615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.736700845366334,
+      "learning_rate": 3.276032110357868e-05,
+      "loss": 1.3945,
+      "step": 9616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.384287788692978,
+      "learning_rate": 3.275635646214652e-05,
+      "loss": 1.9297,
+      "step": 9617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.536036755868622,
+      "learning_rate": 3.275239160485812e-05,
+      "loss": 2.4834,
+      "step": 9618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.833825674090621,
+      "learning_rate": 3.2748426531823816e-05,
+      "loss": 0.6467,
+      "step": 9619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9554545433634831,
+      "learning_rate": 3.2744461243153945e-05,
+      "loss": 0.7402,
+      "step": 9620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.168591549056812,
+      "learning_rate": 3.2740495738958884e-05,
+      "loss": 0.4478,
+      "step": 9621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6459046802262667,
+      "learning_rate": 3.273653001934897e-05,
+      "loss": 1.1035,
+      "step": 9622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.423295618474755,
+      "learning_rate": 3.273256408443458e-05,
+      "loss": 1.1338,
+      "step": 9623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.448509466593016,
+      "learning_rate": 3.272859793432608e-05,
+      "loss": 1.7773,
+      "step": 9624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4382820830974725,
+      "learning_rate": 3.272463156913384e-05,
+      "loss": 0.835,
+      "step": 9625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.281671631364223,
+      "learning_rate": 3.272066498896825e-05,
+      "loss": 0.6904,
+      "step": 9626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3024314053369186,
+      "learning_rate": 3.27166981939397e-05,
+      "loss": 1.2979,
+      "step": 9627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.137539294016017,
+      "learning_rate": 3.271273118415858e-05,
+      "loss": 1.2803,
+      "step": 9628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.235506487722406,
+      "learning_rate": 3.270876395973528e-05,
+      "loss": 1.0625,
+      "step": 9629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.278483241501991,
+      "learning_rate": 3.270479652078023e-05,
+      "loss": 0.6418,
+      "step": 9630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5954912155140866,
+      "learning_rate": 3.270082886740383e-05,
+      "loss": 1.0967,
+      "step": 9631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.607819419015125,
+      "learning_rate": 3.26968609997165e-05,
+      "loss": 1.1299,
+      "step": 9632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2184384389657974,
+      "learning_rate": 3.269289291782865e-05,
+      "loss": 1.2158,
+      "step": 9633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9965931303388427,
+      "learning_rate": 3.2688924621850726e-05,
+      "loss": 1.1279,
+      "step": 9634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.47036841286853,
+      "learning_rate": 3.2684956111893163e-05,
+      "loss": 1.0898,
+      "step": 9635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0523369762930344,
+      "learning_rate": 3.2680987388066395e-05,
+      "loss": 0.5786,
+      "step": 9636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6390400480734835,
+      "learning_rate": 3.267701845048088e-05,
+      "loss": 1.7754,
+      "step": 9637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.611197625284424,
+      "learning_rate": 3.267304929924707e-05,
+      "loss": 1.375,
+      "step": 9638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9290527073719854,
+      "learning_rate": 3.266907993447541e-05,
+      "loss": 1.2852,
+      "step": 9639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.674668040132915,
+      "learning_rate": 3.266511035627638e-05,
+      "loss": 0.8955,
+      "step": 9640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5139211250987517,
+      "learning_rate": 3.266114056476046e-05,
+      "loss": 0.9387,
+      "step": 9641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0927076028574803,
+      "learning_rate": 3.2657170560038114e-05,
+      "loss": 1.0332,
+      "step": 9642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.552885611897089,
+      "learning_rate": 3.2653200342219824e-05,
+      "loss": 1.4102,
+      "step": 9643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.405046663935604,
+      "learning_rate": 3.264922991141608e-05,
+      "loss": 1.2578,
+      "step": 9644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9370327841747668,
+      "learning_rate": 3.264525926773739e-05,
+      "loss": 1.3105,
+      "step": 9645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9480316762687315,
+      "learning_rate": 3.2641288411294245e-05,
+      "loss": 1.4355,
+      "step": 9646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6338955148514052,
+      "learning_rate": 3.263731734219715e-05,
+      "loss": 0.3839,
+      "step": 9647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.369462287734165,
+      "learning_rate": 3.2633346060556626e-05,
+      "loss": 1.1235,
+      "step": 9648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.423483222034491,
+      "learning_rate": 3.262937456648319e-05,
+      "loss": 0.725,
+      "step": 9649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6091695619716995,
+      "learning_rate": 3.2625402860087365e-05,
+      "loss": 1.2256,
+      "step": 9650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5727568253713984,
+      "learning_rate": 3.2621430941479683e-05,
+      "loss": 0.7036,
+      "step": 9651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4754180509741572,
+      "learning_rate": 3.2617458810770676e-05,
+      "loss": 1.1625,
+      "step": 9652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9189088463655042,
+      "learning_rate": 3.261348646807089e-05,
+      "loss": 0.8972,
+      "step": 9653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.366167143770809,
+      "learning_rate": 3.260951391349088e-05,
+      "loss": 0.9961,
+      "step": 9654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.62304192411015,
+      "learning_rate": 3.26055411471412e-05,
+      "loss": 0.7002,
+      "step": 9655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.889827187883858,
+      "learning_rate": 3.2601568169132404e-05,
+      "loss": 1.4805,
+      "step": 9656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.067331838722677,
+      "learning_rate": 3.2597594979575055e-05,
+      "loss": 1.2852,
+      "step": 9657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.713005180538148,
+      "learning_rate": 3.259362157857972e-05,
+      "loss": 0.8701,
+      "step": 9658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8455129162744206,
+      "learning_rate": 3.2589647966257015e-05,
+      "loss": 0.6318,
+      "step": 9659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6021111170494646,
+      "learning_rate": 3.258567414271748e-05,
+      "loss": 1.4199,
+      "step": 9660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.978148550291933,
+      "learning_rate": 3.258170010807173e-05,
+      "loss": 1.6758,
+      "step": 9661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.37637329418773,
+      "learning_rate": 3.257772586243034e-05,
+      "loss": 0.8643,
+      "step": 9662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2997262086862023,
+      "learning_rate": 3.257375140590394e-05,
+      "loss": 1.1094,
+      "step": 9663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.312757903952608,
+      "learning_rate": 3.256977673860311e-05,
+      "loss": 1.0889,
+      "step": 9664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.859210736367812,
+      "learning_rate": 3.256580186063848e-05,
+      "loss": 0.9229,
+      "step": 9665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.698270109337977,
+      "learning_rate": 3.256182677212066e-05,
+      "loss": 1.3359,
+      "step": 9666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.36667720720693,
+      "learning_rate": 3.2557851473160276e-05,
+      "loss": 1.0244,
+      "step": 9667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.402433304423798,
+      "learning_rate": 3.2553875963867974e-05,
+      "loss": 1.2988,
+      "step": 9668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.225406922409112,
+      "learning_rate": 3.254990024435437e-05,
+      "loss": 0.8311,
+      "step": 9669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.22083101162685,
+      "learning_rate": 3.254592431473012e-05,
+      "loss": 1.0938,
+      "step": 9670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5690238528215277,
+      "learning_rate": 3.254194817510586e-05,
+      "loss": 1.0762,
+      "step": 9671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.258357421164714,
+      "learning_rate": 3.2537971825592266e-05,
+      "loss": 0.9175,
+      "step": 9672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.416271874639284,
+      "learning_rate": 3.253399526629998e-05,
+      "loss": 1.3574,
+      "step": 9673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.945235285318098,
+      "learning_rate": 3.2530018497339675e-05,
+      "loss": 1.0332,
+      "step": 9674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0237524376205327,
+      "learning_rate": 3.252604151882202e-05,
+      "loss": 0.9043,
+      "step": 9675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4113965174178817,
+      "learning_rate": 3.252206433085768e-05,
+      "loss": 1.2227,
+      "step": 9676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3067240364827044,
+      "learning_rate": 3.251808693355736e-05,
+      "loss": 0.9561,
+      "step": 9677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6547783758046344,
+      "learning_rate": 3.251410932703175e-05,
+      "loss": 0.7441,
+      "step": 9678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.814661645473321,
+      "learning_rate": 3.251013151139153e-05,
+      "loss": 1.5,
+      "step": 9679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.1009455683155425,
+      "learning_rate": 3.2506153486747404e-05,
+      "loss": 1.2021,
+      "step": 9680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.03637919951456,
+      "learning_rate": 3.2502175253210085e-05,
+      "loss": 1.3105,
+      "step": 9681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.67010091575307,
+      "learning_rate": 3.2498196810890284e-05,
+      "loss": 0.5493,
+      "step": 9682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.697645863574615,
+      "learning_rate": 3.2494218159898727e-05,
+      "loss": 0.6572,
+      "step": 9683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7963932410953927,
+      "learning_rate": 3.2490239300346116e-05,
+      "loss": 0.4629,
+      "step": 9684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.930586261947219,
+      "learning_rate": 3.2486260232343196e-05,
+      "loss": 1.0728,
+      "step": 9685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0500873553164074,
+      "learning_rate": 3.2482280956000705e-05,
+      "loss": 1.0049,
+      "step": 9686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8411762335484863,
+      "learning_rate": 3.247830147142939e-05,
+      "loss": 0.7573,
+      "step": 9687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9926456298640836,
+      "learning_rate": 3.2474321778739966e-05,
+      "loss": 1.4082,
+      "step": 9688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.708789747053696,
+      "learning_rate": 3.247034187804323e-05,
+      "loss": 1.1963,
+      "step": 9689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.936161679359093,
+      "learning_rate": 3.246636176944991e-05,
+      "loss": 2.2695,
+      "step": 9690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.10202030675502,
+      "learning_rate": 3.246238145307079e-05,
+      "loss": 1.2266,
+      "step": 9691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1432610072704206,
+      "learning_rate": 3.245840092901662e-05,
+      "loss": 1.0977,
+      "step": 9692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2327266561825314,
+      "learning_rate": 3.24544201973982e-05,
+      "loss": 1.2715,
+      "step": 9693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.155474425276748,
+      "learning_rate": 3.2450439258326296e-05,
+      "loss": 1.0938,
+      "step": 9694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.900321798145993,
+      "learning_rate": 3.24464581119117e-05,
+      "loss": 1.1221,
+      "step": 9695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.093711407241992,
+      "learning_rate": 3.244247675826521e-05,
+      "loss": 1.3955,
+      "step": 9696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9120235652720883,
+      "learning_rate": 3.243849519749762e-05,
+      "loss": 0.9268,
+      "step": 9697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7154793341006362,
+      "learning_rate": 3.2434513429719734e-05,
+      "loss": 0.9326,
+      "step": 9698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.822005691686211,
+      "learning_rate": 3.243053145504237e-05,
+      "loss": 1.1836,
+      "step": 9699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.203850491868053,
+      "learning_rate": 3.242654927357634e-05,
+      "loss": 1.4258,
+      "step": 9700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.287695086242418,
+      "learning_rate": 3.2422566885432465e-05,
+      "loss": 1.9258,
+      "step": 9701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9074343144817596,
+      "learning_rate": 3.241858429072158e-05,
+      "loss": 1.0913,
+      "step": 9702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.009594240122057,
+      "learning_rate": 3.2414601489554505e-05,
+      "loss": 1.1738,
+      "step": 9703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.9884062532489635,
+      "learning_rate": 3.24106184820421e-05,
+      "loss": 1.375,
+      "step": 9704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7686967762810824,
+      "learning_rate": 3.240663526829519e-05,
+      "loss": 0.7056,
+      "step": 9705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.498676834469511,
+      "learning_rate": 3.240265184842464e-05,
+      "loss": 0.9639,
+      "step": 9706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3906032586199895,
+      "learning_rate": 3.23986682225413e-05,
+      "loss": 0.7153,
+      "step": 9707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.763484751698455,
+      "learning_rate": 3.239468439075604e-05,
+      "loss": 0.7224,
+      "step": 9708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.879887272798715,
+      "learning_rate": 3.239070035317971e-05,
+      "loss": 0.9589,
+      "step": 9709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9444337540758347,
+      "learning_rate": 3.238671610992322e-05,
+      "loss": 1.248,
+      "step": 9710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8564007408381515,
+      "learning_rate": 3.2382731661097413e-05,
+      "loss": 1.2979,
+      "step": 9711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4869894989491415,
+      "learning_rate": 3.23787470068132e-05,
+      "loss": 0.7715,
+      "step": 9712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.244593163820414,
+      "learning_rate": 3.237476214718145e-05,
+      "loss": 1.2734,
+      "step": 9713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.440799827028792,
+      "learning_rate": 3.237077708231308e-05,
+      "loss": 1.2324,
+      "step": 9714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.051532022897282,
+      "learning_rate": 3.2366791812318974e-05,
+      "loss": 0.9355,
+      "step": 9715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.529846838585862,
+      "learning_rate": 3.2362806337310066e-05,
+      "loss": 0.9648,
+      "step": 9716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2821707761554757,
+      "learning_rate": 3.235882065739724e-05,
+      "loss": 1.4707,
+      "step": 9717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.253034029538133,
+      "learning_rate": 3.235483477269144e-05,
+      "loss": 0.7148,
+      "step": 9718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.869273723101935,
+      "learning_rate": 3.2350848683303586e-05,
+      "loss": 1.2832,
+      "step": 9719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9568195141230924,
+      "learning_rate": 3.23468623893446e-05,
+      "loss": 1.0557,
+      "step": 9720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.9139625527022845,
+      "learning_rate": 3.2342875890925425e-05,
+      "loss": 0.8955,
+      "step": 9721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.828240523842489,
+      "learning_rate": 3.2338889188157e-05,
+      "loss": 1.4258,
+      "step": 9722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4819514131178386,
+      "learning_rate": 3.233490228115028e-05,
+      "loss": 0.7935,
+      "step": 9723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.253095473156406,
+      "learning_rate": 3.2330915170016216e-05,
+      "loss": 1.7363,
+      "step": 9724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.62538330964131,
+      "learning_rate": 3.232692785486577e-05,
+      "loss": 0.7959,
+      "step": 9725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1198923713508897,
+      "learning_rate": 3.2322940335809896e-05,
+      "loss": 1.6426,
+      "step": 9726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5552576447743405,
+      "learning_rate": 3.2318952612959575e-05,
+      "loss": 0.5971,
+      "step": 9727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.257962815729977,
+      "learning_rate": 3.231496468642578e-05,
+      "loss": 0.4813,
+      "step": 9728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4511888694668587,
+      "learning_rate": 3.2310976556319514e-05,
+      "loss": 1.0469,
+      "step": 9729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.7307466398727,
+      "learning_rate": 3.230698822275173e-05,
+      "loss": 1.3262,
+      "step": 9730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3892231517543507,
+      "learning_rate": 3.230299968583344e-05,
+      "loss": 1.0078,
+      "step": 9731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9955918987187546,
+      "learning_rate": 3.229901094567565e-05,
+      "loss": 1.4219,
+      "step": 9732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.887675933291313,
+      "learning_rate": 3.229502200238935e-05,
+      "loss": 1.1582,
+      "step": 9733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.331739755684267,
+      "learning_rate": 3.2291032856085555e-05,
+      "loss": 0.9219,
+      "step": 9734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7820650593665426,
+      "learning_rate": 3.228704350687529e-05,
+      "loss": 1.0996,
+      "step": 9735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.004644137112752,
+      "learning_rate": 3.2283053954869566e-05,
+      "loss": 1.1055,
+      "step": 9736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.096432234611662,
+      "learning_rate": 3.227906420017942e-05,
+      "loss": 1.4883,
+      "step": 9737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.218215343363044,
+      "learning_rate": 3.2275074242915884e-05,
+      "loss": 1.6289,
+      "step": 9738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.470763551917519,
+      "learning_rate": 3.2271084083189985e-05,
+      "loss": 1.8105,
+      "step": 9739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.09528397801614,
+      "learning_rate": 3.226709372111279e-05,
+      "loss": 0.9863,
+      "step": 9740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.595018106012096,
+      "learning_rate": 3.2263103156795313e-05,
+      "loss": 1.0762,
+      "step": 9741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.9391490258925765,
+      "learning_rate": 3.2259112390348646e-05,
+      "loss": 1.1895,
+      "step": 9742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7046655223180496,
+      "learning_rate": 3.225512142188383e-05,
+      "loss": 1.4922,
+      "step": 9743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5649313208474145,
+      "learning_rate": 3.225113025151194e-05,
+      "loss": 1.0859,
+      "step": 9744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4220850699594774,
+      "learning_rate": 3.2247138879344045e-05,
+      "loss": 0.8584,
+      "step": 9745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4998190814358723,
+      "learning_rate": 3.224314730549123e-05,
+      "loss": 1.1641,
+      "step": 9746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.644908162561606,
+      "learning_rate": 3.223915553006458e-05,
+      "loss": 1.2715,
+      "step": 9747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.465511852339294,
+      "learning_rate": 3.223516355317517e-05,
+      "loss": 1.0957,
+      "step": 9748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.212931597868097,
+      "learning_rate": 3.2231171374934104e-05,
+      "loss": 1.043,
+      "step": 9749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.5863858317370765,
+      "learning_rate": 3.222717899545248e-05,
+      "loss": 0.6265,
+      "step": 9750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.038822366154109,
+      "learning_rate": 3.2223186414841415e-05,
+      "loss": 0.9482,
+      "step": 9751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.658898055183468,
+      "learning_rate": 3.221919363321201e-05,
+      "loss": 1.416,
+      "step": 9752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.142609504741153,
+      "learning_rate": 3.221520065067538e-05,
+      "loss": 0.7595,
+      "step": 9753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4839909773840065,
+      "learning_rate": 3.221120746734266e-05,
+      "loss": 1.4531,
+      "step": 9754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.044808337313553,
+      "learning_rate": 3.220721408332497e-05,
+      "loss": 1.1035,
+      "step": 9755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.331167755887397,
+      "learning_rate": 3.2203220498733445e-05,
+      "loss": 0.8174,
+      "step": 9756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.797326877446166,
+      "learning_rate": 3.2199226713679224e-05,
+      "loss": 0.7549,
+      "step": 9757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.548562906542077,
+      "learning_rate": 3.219523272827345e-05,
+      "loss": 1.334,
+      "step": 9758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.082724967364658,
+      "learning_rate": 3.219123854262729e-05,
+      "loss": 1.0859,
+      "step": 9759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0555397267920683,
+      "learning_rate": 3.2187244156851883e-05,
+      "loss": 0.741,
+      "step": 9760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.396763717674184,
+      "learning_rate": 3.21832495710584e-05,
+      "loss": 0.6636,
+      "step": 9761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0972188902999984,
+      "learning_rate": 3.2179254785358005e-05,
+      "loss": 0.8032,
+      "step": 9762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.16708419507361,
+      "learning_rate": 3.217525979986187e-05,
+      "loss": 1.3652,
+      "step": 9763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1086957049658346,
+      "learning_rate": 3.217126461468117e-05,
+      "loss": 1.0977,
+      "step": 9764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.141973808420512,
+      "learning_rate": 3.21672692299271e-05,
+      "loss": 0.5688,
+      "step": 9765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7225247191260618,
+      "learning_rate": 3.216327364571085e-05,
+      "loss": 1.1172,
+      "step": 9766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8430823915332435,
+      "learning_rate": 3.215927786214361e-05,
+      "loss": 0.917,
+      "step": 9767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.164532890733131,
+      "learning_rate": 3.2155281879336576e-05,
+      "loss": 0.809,
+      "step": 9768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.0694295293626315,
+      "learning_rate": 3.215128569740096e-05,
+      "loss": 1.0938,
+      "step": 9769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.236568820225915,
+      "learning_rate": 3.214728931644798e-05,
+      "loss": 0.6157,
+      "step": 9770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.948061199262106,
+      "learning_rate": 3.214329273658885e-05,
+      "loss": 1.2241,
+      "step": 9771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.655373714864235,
+      "learning_rate": 3.213929595793479e-05,
+      "loss": 1.1089,
+      "step": 9772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.684101801895675,
+      "learning_rate": 3.213529898059702e-05,
+      "loss": 1.2002,
+      "step": 9773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.027597803100079,
+      "learning_rate": 3.2131301804686787e-05,
+      "loss": 1.0586,
+      "step": 9774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0710331652171683,
+      "learning_rate": 3.2127304430315333e-05,
+      "loss": 0.4485,
+      "step": 9775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6131971314353035,
+      "learning_rate": 3.21233068575939e-05,
+      "loss": 1.0264,
+      "step": 9776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.436102344154504,
+      "learning_rate": 3.211930908663373e-05,
+      "loss": 1.4254,
+      "step": 9777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1273320808889125,
+      "learning_rate": 3.2115311117546087e-05,
+      "loss": 1.0273,
+      "step": 9778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.810552476159153,
+      "learning_rate": 3.211131295044224e-05,
+      "loss": 1.2433,
+      "step": 9779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.226897133462532,
+      "learning_rate": 3.2107314585433444e-05,
+      "loss": 1.8262,
+      "step": 9780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.1770695839114165,
+      "learning_rate": 3.210331602263098e-05,
+      "loss": 1.71,
+      "step": 9781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.782865695009313,
+      "learning_rate": 3.209931726214611e-05,
+      "loss": 0.4838,
+      "step": 9782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.747549447024888,
+      "learning_rate": 3.209531830409014e-05,
+      "loss": 1.1096,
+      "step": 9783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.600722168617756,
+      "learning_rate": 3.209131914857435e-05,
+      "loss": 1.3379,
+      "step": 9784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7996143145336454,
+      "learning_rate": 3.2087319795710034e-05,
+      "loss": 0.6528,
+      "step": 9785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2603623930103764,
+      "learning_rate": 3.20833202456085e-05,
+      "loss": 0.9795,
+      "step": 9786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7491452882908156,
+      "learning_rate": 3.207932049838104e-05,
+      "loss": 1.541,
+      "step": 9787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8349426037665015,
+      "learning_rate": 3.207532055413897e-05,
+      "loss": 1.25,
+      "step": 9788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.171445217211334,
+      "learning_rate": 3.207132041299362e-05,
+      "loss": 1.2734,
+      "step": 9789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.7851185201983375,
+      "learning_rate": 3.206732007505629e-05,
+      "loss": 1.0938,
+      "step": 9790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9827652056068528,
+      "learning_rate": 3.206331954043833e-05,
+      "loss": 1.3164,
+      "step": 9791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.842389709445588,
+      "learning_rate": 3.205931880925105e-05,
+      "loss": 1.2812,
+      "step": 9792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5843263908318126,
+      "learning_rate": 3.2055317881605814e-05,
+      "loss": 0.3792,
+      "step": 9793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.114494384498383,
+      "learning_rate": 3.205131675761395e-05,
+      "loss": 0.4053,
+      "step": 9794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9435302045873093,
+      "learning_rate": 3.204731543738681e-05,
+      "loss": 0.8945,
+      "step": 9795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.241934732152388,
+      "learning_rate": 3.2043313921035743e-05,
+      "loss": 0.7793,
+      "step": 9796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7527460851930337,
+      "learning_rate": 3.2039312208672126e-05,
+      "loss": 1.1016,
+      "step": 9797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5741380785977213,
+      "learning_rate": 3.203531030040731e-05,
+      "loss": 0.8994,
+      "step": 9798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8030806484751023,
+      "learning_rate": 3.203130819635268e-05,
+      "loss": 0.9016,
+      "step": 9799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.395101659542095,
+      "learning_rate": 3.2027305896619595e-05,
+      "loss": 0.9727,
+      "step": 9800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8827810725124676,
+      "learning_rate": 3.202330340131946e-05,
+      "loss": 1.3057,
+      "step": 9801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2057142260937574,
+      "learning_rate": 3.201930071056364e-05,
+      "loss": 1.1367,
+      "step": 9802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.122374090227646,
+      "learning_rate": 3.201529782446355e-05,
+      "loss": 1.2578,
+      "step": 9803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0847944349500107,
+      "learning_rate": 3.2011294743130566e-05,
+      "loss": 0.9434,
+      "step": 9804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.512274839266491,
+      "learning_rate": 3.200729146667611e-05,
+      "loss": 0.7422,
+      "step": 9805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.817897195650531,
+      "learning_rate": 3.200328799521159e-05,
+      "loss": 1.5684,
+      "step": 9806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.446857878136685,
+      "learning_rate": 3.19992843288484e-05,
+      "loss": 1.2852,
+      "step": 9807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.274921889588574,
+      "learning_rate": 3.199528046769799e-05,
+      "loss": 0.4045,
+      "step": 9808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.149085706546966,
+      "learning_rate": 3.199127641187177e-05,
+      "loss": 0.4414,
+      "step": 9809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5264301329071386,
+      "learning_rate": 3.198727216148117e-05,
+      "loss": 1.4707,
+      "step": 9810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3733864372541595,
+      "learning_rate": 3.198326771663763e-05,
+      "loss": 1.1846,
+      "step": 9811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.810537543688905,
+      "learning_rate": 3.19792630774526e-05,
+      "loss": 0.7031,
+      "step": 9812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.146922953945722,
+      "learning_rate": 3.197525824403751e-05,
+      "loss": 1.0918,
+      "step": 9813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.882323515651384,
+      "learning_rate": 3.197125321650383e-05,
+      "loss": 1.0586,
+      "step": 9814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.160515122796312,
+      "learning_rate": 3.1967247994963e-05,
+      "loss": 1.918,
+      "step": 9815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.840914149025151,
+      "learning_rate": 3.1963242579526506e-05,
+      "loss": 2.0098,
+      "step": 9816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7598973096886688,
+      "learning_rate": 3.195923697030581e-05,
+      "loss": 1.0908,
+      "step": 9817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8053390845926054,
+      "learning_rate": 3.195523116741237e-05,
+      "loss": 0.6499,
+      "step": 9818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.30714342546688,
+      "learning_rate": 3.1951225170957685e-05,
+      "loss": 0.9355,
+      "step": 9819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.178069698221685,
+      "learning_rate": 3.194721898105323e-05,
+      "loss": 1.3086,
+      "step": 9820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.723502097258654,
+      "learning_rate": 3.194321259781049e-05,
+      "loss": 1.6035,
+      "step": 9821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8155476827786687,
+      "learning_rate": 3.193920602134099e-05,
+      "loss": 0.3021,
+      "step": 9822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.894997748339474,
+      "learning_rate": 3.19351992517562e-05,
+      "loss": 0.6895,
+      "step": 9823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.084866031105825,
+      "learning_rate": 3.193119228916764e-05,
+      "loss": 0.9297,
+      "step": 9824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.380223809036277,
+      "learning_rate": 3.1927185133686824e-05,
+      "loss": 1.5859,
+      "step": 9825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.31713757265238,
+      "learning_rate": 3.192317778542526e-05,
+      "loss": 1.0078,
+      "step": 9826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.023081228537541,
+      "learning_rate": 3.191917024449448e-05,
+      "loss": 0.8018,
+      "step": 9827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.085103228792704,
+      "learning_rate": 3.191516251100601e-05,
+      "loss": 1.1953,
+      "step": 9828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8993123206267413,
+      "learning_rate": 3.1911154585071374e-05,
+      "loss": 1.4512,
+      "step": 9829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.622886442616805,
+      "learning_rate": 3.1907146466802125e-05,
+      "loss": 0.7705,
+      "step": 9830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.331697849008794,
+      "learning_rate": 3.19031381563098e-05,
+      "loss": 1.8516,
+      "step": 9831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.904121246127965,
+      "learning_rate": 3.189912965370596e-05,
+      "loss": 0.6602,
+      "step": 9832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.109665651554043,
+      "learning_rate": 3.1895120959102135e-05,
+      "loss": 1.3906,
+      "step": 9833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8335211261522426,
+      "learning_rate": 3.189111207260991e-05,
+      "loss": 1.1475,
+      "step": 9834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.0241666415938955,
+      "learning_rate": 3.188710299434085e-05,
+      "loss": 1.2383,
+      "step": 9835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.22320189509484,
+      "learning_rate": 3.18830937244065e-05,
+      "loss": 0.8174,
+      "step": 9836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.061090649444176,
+      "learning_rate": 3.1879084262918465e-05,
+      "loss": 1.0449,
+      "step": 9837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4417292755050464,
+      "learning_rate": 3.187507460998831e-05,
+      "loss": 1.0342,
+      "step": 9838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8977930221306156,
+      "learning_rate": 3.1871064765727626e-05,
+      "loss": 0.7578,
+      "step": 9839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.972951957812153,
+      "learning_rate": 3.1867054730248015e-05,
+      "loss": 1.5273,
+      "step": 9840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.811634942768499,
+      "learning_rate": 3.1863044503661054e-05,
+      "loss": 1.8457,
+      "step": 9841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9361283366663318,
+      "learning_rate": 3.185903408607836e-05,
+      "loss": 0.7148,
+      "step": 9842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.405066161802054,
+      "learning_rate": 3.185502347761155e-05,
+      "loss": 1.0229,
+      "step": 9843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.01074570852902,
+      "learning_rate": 3.1851012678372224e-05,
+      "loss": 1.4941,
+      "step": 9844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.99132387385915,
+      "learning_rate": 3.1847001688472e-05,
+      "loss": 0.8906,
+      "step": 9845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3172081650653125,
+      "learning_rate": 3.184299050802252e-05,
+      "loss": 0.7402,
+      "step": 9846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9782733956318412,
+      "learning_rate": 3.183897913713538e-05,
+      "loss": 1.3574,
+      "step": 9847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.4533199942191715,
+      "learning_rate": 3.183496757592225e-05,
+      "loss": 1.6484,
+      "step": 9848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5292993258313414,
+      "learning_rate": 3.1830955824494747e-05,
+      "loss": 0.5811,
+      "step": 9849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.219951995287304,
+      "learning_rate": 3.182694388296453e-05,
+      "loss": 0.7632,
+      "step": 9850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.375636711813646,
+      "learning_rate": 3.182293175144323e-05,
+      "loss": 1.248,
+      "step": 9851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.026292147019747,
+      "learning_rate": 3.1818919430042524e-05,
+      "loss": 1.0518,
+      "step": 9852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2572593310810474,
+      "learning_rate": 3.181490691887407e-05,
+      "loss": 1.0449,
+      "step": 9853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7715468288748064,
+      "learning_rate": 3.181089421804954e-05,
+      "loss": 0.8354,
+      "step": 9854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.536774198960387,
+      "learning_rate": 3.180688132768058e-05,
+      "loss": 1.3936,
+      "step": 9855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9238995560537986,
+      "learning_rate": 3.180286824787889e-05,
+      "loss": 1.1895,
+      "step": 9856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.899897698538227,
+      "learning_rate": 3.179885497875615e-05,
+      "loss": 1.125,
+      "step": 9857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.996265789399381,
+      "learning_rate": 3.179484152042404e-05,
+      "loss": 0.9578,
+      "step": 9858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8946781919095597,
+      "learning_rate": 3.1790827872994264e-05,
+      "loss": 1.3457,
+      "step": 9859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4412184498081766,
+      "learning_rate": 3.17868140365785e-05,
+      "loss": 0.9355,
+      "step": 9860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.9667006564459735,
+      "learning_rate": 3.1782800011288466e-05,
+      "loss": 1.1582,
+      "step": 9861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.616185224515322,
+      "learning_rate": 3.1778785797235874e-05,
+      "loss": 0.7466,
+      "step": 9862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.240632265470209,
+      "learning_rate": 3.177477139453243e-05,
+      "loss": 1.5938,
+      "step": 9863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.894808489537965,
+      "learning_rate": 3.177075680328986e-05,
+      "loss": 1.4902,
+      "step": 9864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0174117255395125,
+      "learning_rate": 3.176674202361989e-05,
+      "loss": 1.3301,
+      "step": 9865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3587183764935045,
+      "learning_rate": 3.176272705563423e-05,
+      "loss": 1.082,
+      "step": 9866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1474391215696,
+      "learning_rate": 3.175871189944464e-05,
+      "loss": 1.3828,
+      "step": 9867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8952471137659488,
+      "learning_rate": 3.175469655516284e-05,
+      "loss": 0.8608,
+      "step": 9868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5429302159918863,
+      "learning_rate": 3.1750681022900595e-05,
+      "loss": 1.334,
+      "step": 9869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9651671005135094,
+      "learning_rate": 3.1746665302769626e-05,
+      "loss": 1.1345,
+      "step": 9870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2202907102134413,
+      "learning_rate": 3.174264939488173e-05,
+      "loss": 0.6841,
+      "step": 9871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9188156022535057,
+      "learning_rate": 3.173863329934864e-05,
+      "loss": 1.2373,
+      "step": 9872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2182362996046634,
+      "learning_rate": 3.1734617016282134e-05,
+      "loss": 1.2891,
+      "step": 9873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1322796613625177,
+      "learning_rate": 3.1730600545793964e-05,
+      "loss": 0.7305,
+      "step": 9874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.411462859334193,
+      "learning_rate": 3.172658388799593e-05,
+      "loss": 0.8149,
+      "step": 9875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8809344156889274,
+      "learning_rate": 3.172256704299981e-05,
+      "loss": 0.936,
+      "step": 9876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.094602833348246,
+      "learning_rate": 3.171855001091737e-05,
+      "loss": 1.1172,
+      "step": 9877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.896444370561992,
+      "learning_rate": 3.171453279186044e-05,
+      "loss": 0.7964,
+      "step": 9878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.135289392826911,
+      "learning_rate": 3.171051538594077e-05,
+      "loss": 0.8389,
+      "step": 9879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1206466737556466,
+      "learning_rate": 3.1706497793270205e-05,
+      "loss": 1.0889,
+      "step": 9880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.860822750197531,
+      "learning_rate": 3.1702480013960535e-05,
+      "loss": 1.2207,
+      "step": 9881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.298485950034523,
+      "learning_rate": 3.169846204812357e-05,
+      "loss": 0.8677,
+      "step": 9882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3023521883486495,
+      "learning_rate": 3.169444389587114e-05,
+      "loss": 1.2207,
+      "step": 9883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.165592674984328,
+      "learning_rate": 3.169042555731506e-05,
+      "loss": 1.666,
+      "step": 9884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5019632260359135,
+      "learning_rate": 3.168640703256716e-05,
+      "loss": 1.5312,
+      "step": 9885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5141183831527973,
+      "learning_rate": 3.168238832173927e-05,
+      "loss": 1.2224,
+      "step": 9886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.008328826542235,
+      "learning_rate": 3.167836942494323e-05,
+      "loss": 1.0391,
+      "step": 9887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2221629002105345,
+      "learning_rate": 3.1674350342290906e-05,
+      "loss": 0.9043,
+      "step": 9888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.258909282461963,
+      "learning_rate": 3.167033107389411e-05,
+      "loss": 1.2422,
+      "step": 9889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3825803329052113,
+      "learning_rate": 3.166631161986472e-05,
+      "loss": 1.146,
+      "step": 9890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.916668991814549,
+      "learning_rate": 3.16622919803146e-05,
+      "loss": 0.8657,
+      "step": 9891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.549906870129923,
+      "learning_rate": 3.16582721553556e-05,
+      "loss": 1.3223,
+      "step": 9892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.4833767091758485,
+      "learning_rate": 3.165425214509959e-05,
+      "loss": 1.5176,
+      "step": 9893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9239429420250724,
+      "learning_rate": 3.165023194965846e-05,
+      "loss": 0.5906,
+      "step": 9894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.611916309072524,
+      "learning_rate": 3.164621156914408e-05,
+      "loss": 1.2173,
+      "step": 9895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.419916160644484,
+      "learning_rate": 3.164219100366834e-05,
+      "loss": 0.5132,
+      "step": 9896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.288357362428861,
+      "learning_rate": 3.163817025334313e-05,
+      "loss": 1.0049,
+      "step": 9897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.566961266178227,
+      "learning_rate": 3.163414931828034e-05,
+      "loss": 1.1777,
+      "step": 9898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.257031051408621,
+      "learning_rate": 3.163012819859188e-05,
+      "loss": 0.8618,
+      "step": 9899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7631272065698096,
+      "learning_rate": 3.1626106894389646e-05,
+      "loss": 0.9785,
+      "step": 9900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.467075274172907,
+      "learning_rate": 3.162208540578556e-05,
+      "loss": 1.3457,
+      "step": 9901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.261112942806438,
+      "learning_rate": 3.161806373289153e-05,
+      "loss": 1.0469,
+      "step": 9902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.344178835075099,
+      "learning_rate": 3.1614041875819475e-05,
+      "loss": 0.9443,
+      "step": 9903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.039629839189589,
+      "learning_rate": 3.161001983468133e-05,
+      "loss": 1.1787,
+      "step": 9904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5288997694846547,
+      "learning_rate": 3.160599760958903e-05,
+      "loss": 0.8428,
+      "step": 9905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1342121907296825,
+      "learning_rate": 3.16019752006545e-05,
+      "loss": 1.085,
+      "step": 9906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5260206003090384,
+      "learning_rate": 3.159795260798969e-05,
+      "loss": 1.2842,
+      "step": 9907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.155621745293345,
+      "learning_rate": 3.1593929831706545e-05,
+      "loss": 1.002,
+      "step": 9908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.106411415639791,
+      "learning_rate": 3.158990687191702e-05,
+      "loss": 1.3809,
+      "step": 9909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.403111777002592,
+      "learning_rate": 3.1585883728733065e-05,
+      "loss": 0.6885,
+      "step": 9910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.164079120741975,
+      "learning_rate": 3.158186040226665e-05,
+      "loss": 1.0879,
+      "step": 9911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6231463108828565,
+      "learning_rate": 3.157783689262973e-05,
+      "loss": 0.8838,
+      "step": 9912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.103751043714053,
+      "learning_rate": 3.15738131999343e-05,
+      "loss": 0.6885,
+      "step": 9913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4740069468222607,
+      "learning_rate": 3.1569789324292324e-05,
+      "loss": 1.0586,
+      "step": 9914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.302001289016525,
+      "learning_rate": 3.1565765265815784e-05,
+      "loss": 2.4912,
+      "step": 9915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5613697629810543,
+      "learning_rate": 3.1561741024616667e-05,
+      "loss": 1.2266,
+      "step": 9916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.920452087763442,
+      "learning_rate": 3.155771660080696e-05,
+      "loss": 0.835,
+      "step": 9917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7295717967411464,
+      "learning_rate": 3.1553691994498684e-05,
+      "loss": 0.8492,
+      "step": 9918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3153629623188308,
+      "learning_rate": 3.154966720580382e-05,
+      "loss": 1.2148,
+      "step": 9919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.061950411668093,
+      "learning_rate": 3.1545642234834396e-05,
+      "loss": 0.5176,
+      "step": 9920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.349738544772648,
+      "learning_rate": 3.154161708170239e-05,
+      "loss": 0.8535,
+      "step": 9921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.243195256385369,
+      "learning_rate": 3.153759174651986e-05,
+      "loss": 1.1035,
+      "step": 9922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.726046346429912,
+      "learning_rate": 3.153356622939881e-05,
+      "loss": 1.0439,
+      "step": 9923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.112989818059217,
+      "learning_rate": 3.152954053045127e-05,
+      "loss": 1.25,
+      "step": 9924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.193620079667717,
+      "learning_rate": 3.1525514649789284e-05,
+      "loss": 0.3901,
+      "step": 9925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.712893525157062,
+      "learning_rate": 3.1521488587524875e-05,
+      "loss": 0.9404,
+      "step": 9926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.1654747028951755,
+      "learning_rate": 3.151746234377009e-05,
+      "loss": 0.954,
+      "step": 9927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.088509948121585,
+      "learning_rate": 3.1513435918636984e-05,
+      "loss": 1.1631,
+      "step": 9928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.767772574397391,
+      "learning_rate": 3.150940931223761e-05,
+      "loss": 1.0586,
+      "step": 9929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5844718721216693,
+      "learning_rate": 3.150538252468403e-05,
+      "loss": 1.1445,
+      "step": 9930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.260048363092176,
+      "learning_rate": 3.150135555608829e-05,
+      "loss": 1.3086,
+      "step": 9931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.5646852760385626,
+      "learning_rate": 3.1497328406562476e-05,
+      "loss": 1.5898,
+      "step": 9932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.437218143568186,
+      "learning_rate": 3.149330107621867e-05,
+      "loss": 1.4648,
+      "step": 9933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.435228047803738,
+      "learning_rate": 3.148927356516892e-05,
+      "loss": 0.9502,
+      "step": 9934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9262247870872715,
+      "learning_rate": 3.148524587352534e-05,
+      "loss": 0.8564,
+      "step": 9935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.759423293540085,
+      "learning_rate": 3.14812180014e-05,
+      "loss": 0.8928,
+      "step": 9936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2121498065469654,
+      "learning_rate": 3.1477189948905004e-05,
+      "loss": 1.4414,
+      "step": 9937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.268713870251683,
+      "learning_rate": 3.1473161716152454e-05,
+      "loss": 0.9248,
+      "step": 9938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.310159683843785,
+      "learning_rate": 3.146913330325445e-05,
+      "loss": 1.0527,
+      "step": 9939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.295616895254749,
+      "learning_rate": 3.1465104710323085e-05,
+      "loss": 1.1367,
+      "step": 9940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9747655739424244,
+      "learning_rate": 3.14610759374705e-05,
+      "loss": 0.7773,
+      "step": 9941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7725137381045593,
+      "learning_rate": 3.14570469848088e-05,
+      "loss": 1.9336,
+      "step": 9942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.978035778889387,
+      "learning_rate": 3.1453017852450114e-05,
+      "loss": 1.5254,
+      "step": 9943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.789290767583137,
+      "learning_rate": 3.144898854050656e-05,
+      "loss": 0.8376,
+      "step": 9944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2810678278677403,
+      "learning_rate": 3.1444959049090285e-05,
+      "loss": 0.9561,
+      "step": 9945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8223236606003086,
+      "learning_rate": 3.144092937831343e-05,
+      "loss": 1.0996,
+      "step": 9946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.452718080366492,
+      "learning_rate": 3.143689952828812e-05,
+      "loss": 1.1484,
+      "step": 9947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.468498596375582,
+      "learning_rate": 3.143286949912653e-05,
+      "loss": 1.2119,
+      "step": 9948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.07277190758894,
+      "learning_rate": 3.142883929094078e-05,
+      "loss": 0.7461,
+      "step": 9949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.395346452566952,
+      "learning_rate": 3.142480890384307e-05,
+      "loss": 1.228,
+      "step": 9950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5841790527453683,
+      "learning_rate": 3.1420778337945535e-05,
+      "loss": 0.7354,
+      "step": 9951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.345790905012331,
+      "learning_rate": 3.141674759336036e-05,
+      "loss": 1.0479,
+      "step": 9952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.835318234532162,
+      "learning_rate": 3.14127166701997e-05,
+      "loss": 1.6328,
+      "step": 9953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0191080011418263,
+      "learning_rate": 3.1408685568575754e-05,
+      "loss": 1.3301,
+      "step": 9954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.660119628502332,
+      "learning_rate": 3.140465428860069e-05,
+      "loss": 0.7988,
+      "step": 9955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.703466117795032,
+      "learning_rate": 3.140062283038671e-05,
+      "loss": 1.0625,
+      "step": 9956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7351889261979894,
+      "learning_rate": 3.1396591194046004e-05,
+      "loss": 0.6567,
+      "step": 9957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6900504674041708,
+      "learning_rate": 3.139255937969077e-05,
+      "loss": 0.8652,
+      "step": 9958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.771881010614564,
+      "learning_rate": 3.13885273874332e-05,
+      "loss": 1.041,
+      "step": 9959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.613222919714989,
+      "learning_rate": 3.138449521738552e-05,
+      "loss": 1.1484,
+      "step": 9960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.860854496787562,
+      "learning_rate": 3.138046286965995e-05,
+      "loss": 1.1084,
+      "step": 9961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.104109264232652,
+      "learning_rate": 3.137643034436868e-05,
+      "loss": 0.9316,
+      "step": 9962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5179976179673713,
+      "learning_rate": 3.137239764162395e-05,
+      "loss": 1.1719,
+      "step": 9963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.579963629562344,
+      "learning_rate": 3.136836476153799e-05,
+      "loss": 1.125,
+      "step": 9964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.265004469533619,
+      "learning_rate": 3.1364331704223035e-05,
+      "loss": 1.5977,
+      "step": 9965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5916072773236127,
+      "learning_rate": 3.136029846979132e-05,
+      "loss": 1.1514,
+      "step": 9966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2067926018373454,
+      "learning_rate": 3.1356265058355096e-05,
+      "loss": 0.8936,
+      "step": 9967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5253434202616463,
+      "learning_rate": 3.1352231470026584e-05,
+      "loss": 0.9438,
+      "step": 9968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.572269978174636,
+      "learning_rate": 3.134819770491806e-05,
+      "loss": 0.9313,
+      "step": 9969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4200055197975954,
+      "learning_rate": 3.134416376314179e-05,
+      "loss": 1.0957,
+      "step": 9970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.396219824977558,
+      "learning_rate": 3.134012964481002e-05,
+      "loss": 1.3242,
+      "step": 9971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.9836955309920254,
+      "learning_rate": 3.133609535003501e-05,
+      "loss": 1.1831,
+      "step": 9972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3975285600188285,
+      "learning_rate": 3.1332060878929066e-05,
+      "loss": 1.2871,
+      "step": 9973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.923274153188219,
+      "learning_rate": 3.1328026231604436e-05,
+      "loss": 0.8687,
+      "step": 9974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9760171492253282,
+      "learning_rate": 3.1323991408173416e-05,
+      "loss": 0.2324,
+      "step": 9975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.372009510000621,
+      "learning_rate": 3.131995640874828e-05,
+      "loss": 1.1548,
+      "step": 9976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.674858623250447,
+      "learning_rate": 3.131592123344134e-05,
+      "loss": 1.1113,
+      "step": 9977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.181716112919174,
+      "learning_rate": 3.131188588236488e-05,
+      "loss": 1.1855,
+      "step": 9978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.841899296768658,
+      "learning_rate": 3.13078503556312e-05,
+      "loss": 0.6882,
+      "step": 9979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.5993726378933095,
+      "learning_rate": 3.130381465335262e-05,
+      "loss": 1.5508,
+      "step": 9980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.12239686151019,
+      "learning_rate": 3.129977877564144e-05,
+      "loss": 1.3242,
+      "step": 9981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.724283834718654,
+      "learning_rate": 3.129574272260999e-05,
+      "loss": 1.1855,
+      "step": 9982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4441122698743287,
+      "learning_rate": 3.129170649437057e-05,
+      "loss": 1.0391,
+      "step": 9983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.52824873839522,
+      "learning_rate": 3.128767009103553e-05,
+      "loss": 0.8232,
+      "step": 9984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9528257834008254,
+      "learning_rate": 3.1283633512717185e-05,
+      "loss": 1.2812,
+      "step": 9985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.200352090882136,
+      "learning_rate": 3.127959675952788e-05,
+      "loss": 0.939,
+      "step": 9986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.5941380201053565,
+      "learning_rate": 3.1275559831579946e-05,
+      "loss": 0.8867,
+      "step": 9987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.563653267719214,
+      "learning_rate": 3.1271522728985754e-05,
+      "loss": 1.2422,
+      "step": 9988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.94670328889299,
+      "learning_rate": 3.1267485451857624e-05,
+      "loss": 0.8545,
+      "step": 9989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.311049809695583,
+      "learning_rate": 3.1263448000307935e-05,
+      "loss": 0.8838,
+      "step": 9990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.939876361934511,
+      "learning_rate": 3.125941037444902e-05,
+      "loss": 1.7051,
+      "step": 9991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.302683227074555,
+      "learning_rate": 3.1255372574393285e-05,
+      "loss": 0.915,
+      "step": 9992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.506093229113844,
+      "learning_rate": 3.1251334600253066e-05,
+      "loss": 1.4336,
+      "step": 9993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.849878697890179,
+      "learning_rate": 3.124729645214075e-05,
+      "loss": 1.0254,
+      "step": 9994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.964377304999565,
+      "learning_rate": 3.1243258130168714e-05,
+      "loss": 1.5254,
+      "step": 9995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.186016901907081,
+      "learning_rate": 3.123921963444935e-05,
+      "loss": 0.9084,
+      "step": 9996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.549164814852866,
+      "learning_rate": 3.123518096509504e-05,
+      "loss": 1.0288,
+      "step": 9997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.542621306245833,
+      "learning_rate": 3.1231142122218196e-05,
+      "loss": 2.4297,
+      "step": 9998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5595853705685465,
+      "learning_rate": 3.122710310593119e-05,
+      "loss": 0.3405,
+      "step": 9999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0670101664461407,
+      "learning_rate": 3.122306391634644e-05,
+      "loss": 1.0957,
+      "step": 10000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1941251428569633,
+      "learning_rate": 3.1219024553576356e-05,
+      "loss": 0.7295,
+      "step": 10001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.5986088598003825,
+      "learning_rate": 3.121498501773336e-05,
+      "loss": 1.0879,
+      "step": 10002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.972038455483304,
+      "learning_rate": 3.121094530892985e-05,
+      "loss": 0.902,
+      "step": 10003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6463960417993713,
+      "learning_rate": 3.1206905427278264e-05,
+      "loss": 1.2637,
+      "step": 10004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8939017204569293,
+      "learning_rate": 3.120286537289102e-05,
+      "loss": 1.5293,
+      "step": 10005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.030437513582626,
+      "learning_rate": 3.119882514588056e-05,
+      "loss": 0.8398,
+      "step": 10006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.10992764109735,
+      "learning_rate": 3.119478474635933e-05,
+      "loss": 1.0,
+      "step": 10007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.167852615132657,
+      "learning_rate": 3.1190744174439754e-05,
+      "loss": 1.306,
+      "step": 10008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5208668603749174,
+      "learning_rate": 3.118670343023429e-05,
+      "loss": 0.835,
+      "step": 10009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.249651479540492,
+      "learning_rate": 3.118266251385539e-05,
+      "loss": 1.6348,
+      "step": 10010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.780994532135681,
+      "learning_rate": 3.117862142541551e-05,
+      "loss": 1.4395,
+      "step": 10011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.463392674463448,
+      "learning_rate": 3.117458016502711e-05,
+      "loss": 1.1206,
+      "step": 10012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7930814280155514,
+      "learning_rate": 3.117053873280266e-05,
+      "loss": 1.3984,
+      "step": 10013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.670007091329714,
+      "learning_rate": 3.1166497128854615e-05,
+      "loss": 0.7393,
+      "step": 10014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.473679290756556,
+      "learning_rate": 3.116245535329548e-05,
+      "loss": 1.2383,
+      "step": 10015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.9931082882212054,
+      "learning_rate": 3.115841340623772e-05,
+      "loss": 1.1934,
+      "step": 10016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8707321550271865,
+      "learning_rate": 3.115437128779382e-05,
+      "loss": 1.1709,
+      "step": 10017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.261060779782387,
+      "learning_rate": 3.1150328998076264e-05,
+      "loss": 1.5664,
+      "step": 10018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.325080426822117,
+      "learning_rate": 3.1146286537197564e-05,
+      "loss": 1.2188,
+      "step": 10019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2299699502025443,
+      "learning_rate": 3.1142243905270215e-05,
+      "loss": 1.8574,
+      "step": 10020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.877712776844065,
+      "learning_rate": 3.113820110240671e-05,
+      "loss": 1.0479,
+      "step": 10021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.209865843057874,
+      "learning_rate": 3.1134158128719575e-05,
+      "loss": 1.0586,
+      "step": 10022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9710731121227163,
+      "learning_rate": 3.113011498432131e-05,
+      "loss": 0.5215,
+      "step": 10023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3344126543840713,
+      "learning_rate": 3.112607166932444e-05,
+      "loss": 1.2363,
+      "step": 10024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.277613745402733,
+      "learning_rate": 3.1122028183841496e-05,
+      "loss": 0.9138,
+      "step": 10025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6082536020492877,
+      "learning_rate": 3.1117984527985e-05,
+      "loss": 1.0684,
+      "step": 10026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.005593604252408,
+      "learning_rate": 3.1113940701867475e-05,
+      "loss": 0.8252,
+      "step": 10027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8453763916957513,
+      "learning_rate": 3.1109896705601485e-05,
+      "loss": 0.6787,
+      "step": 10028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2497175887149456,
+      "learning_rate": 3.110585253929954e-05,
+      "loss": 0.7939,
+      "step": 10029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.1654959346942455,
+      "learning_rate": 3.110180820307422e-05,
+      "loss": 1.5703,
+      "step": 10030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2087720639088033,
+      "learning_rate": 3.1097763697038054e-05,
+      "loss": 0.6226,
+      "step": 10031
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.215568970543361,
+      "learning_rate": 3.1093719021303606e-05,
+      "loss": 1.1309,
+      "step": 10032
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.13035279571488,
+      "learning_rate": 3.108967417598344e-05,
+      "loss": 0.8008,
+      "step": 10033
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.299518335742513,
+      "learning_rate": 3.1085629161190125e-05,
+      "loss": 1.0938,
+      "step": 10034
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4521546532387464,
+      "learning_rate": 3.108158397703622e-05,
+      "loss": 0.8784,
+      "step": 10035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1867101060615375,
+      "learning_rate": 3.107753862363431e-05,
+      "loss": 0.7817,
+      "step": 10036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.162423620696361,
+      "learning_rate": 3.107349310109698e-05,
+      "loss": 1.5312,
+      "step": 10037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.321723473044732,
+      "learning_rate": 3.10694474095368e-05,
+      "loss": 1.0215,
+      "step": 10038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6372134048449873,
+      "learning_rate": 3.106540154906638e-05,
+      "loss": 1.0254,
+      "step": 10039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.917516904257891,
+      "learning_rate": 3.10613555197983e-05,
+      "loss": 1.5391,
+      "step": 10040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0262244631589463,
+      "learning_rate": 3.105730932184516e-05,
+      "loss": 1.0098,
+      "step": 10041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.385173890754245,
+      "learning_rate": 3.105326295531957e-05,
+      "loss": 1.5215,
+      "step": 10042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9019663425441857,
+      "learning_rate": 3.1049216420334135e-05,
+      "loss": 1.2773,
+      "step": 10043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8574535813987656,
+      "learning_rate": 3.104516971700148e-05,
+      "loss": 1.3564,
+      "step": 10044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.495996711148933,
+      "learning_rate": 3.10411228454342e-05,
+      "loss": 1.9141,
+      "step": 10045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.029201489797419,
+      "learning_rate": 3.1037075805744936e-05,
+      "loss": 1.3164,
+      "step": 10046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3978226028666887,
+      "learning_rate": 3.103302859804631e-05,
+      "loss": 0.5809,
+      "step": 10047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.1566278959871115,
+      "learning_rate": 3.102898122245095e-05,
+      "loss": 0.5417,
+      "step": 10048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.78931201491892,
+      "learning_rate": 3.1024933679071496e-05,
+      "loss": 1.1367,
+      "step": 10049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.014041098437011,
+      "learning_rate": 3.10208859680206e-05,
+      "loss": 0.9912,
+      "step": 10050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.613961320666171,
+      "learning_rate": 3.101683808941089e-05,
+      "loss": 0.5884,
+      "step": 10051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1069779093889855,
+      "learning_rate": 3.101279004335503e-05,
+      "loss": 0.7085,
+      "step": 10052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3526517556808746,
+      "learning_rate": 3.100874182996567e-05,
+      "loss": 1.1855,
+      "step": 10053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4718312217414105,
+      "learning_rate": 3.100469344935547e-05,
+      "loss": 1.1875,
+      "step": 10054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.735102330102073,
+      "learning_rate": 3.10006449016371e-05,
+      "loss": 0.894,
+      "step": 10055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.541492888694809,
+      "learning_rate": 3.099659618692322e-05,
+      "loss": 1.5918,
+      "step": 10056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.928833901186799,
+      "learning_rate": 3.099254730532651e-05,
+      "loss": 0.8965,
+      "step": 10057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.450508714916873,
+      "learning_rate": 3.098849825695965e-05,
+      "loss": 0.2988,
+      "step": 10058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6221405985982402,
+      "learning_rate": 3.098444904193532e-05,
+      "loss": 0.604,
+      "step": 10059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.778304284624508,
+      "learning_rate": 3.098039966036621e-05,
+      "loss": 1.4375,
+      "step": 10060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.388405361575713,
+      "learning_rate": 3.0976350112365006e-05,
+      "loss": 1.6582,
+      "step": 10061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.8542135469815335,
+      "learning_rate": 3.0972300398044426e-05,
+      "loss": 1.3174,
+      "step": 10062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4958933479680483,
+      "learning_rate": 3.0968250517517144e-05,
+      "loss": 0.9319,
+      "step": 10063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.18171897266551,
+      "learning_rate": 3.096420047089589e-05,
+      "loss": 1.0938,
+      "step": 10064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.534189084104872,
+      "learning_rate": 3.096015025829335e-05,
+      "loss": 1.8242,
+      "step": 10065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6033829000371007,
+      "learning_rate": 3.095609987982227e-05,
+      "loss": 0.9609,
+      "step": 10066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0317650948835975,
+      "learning_rate": 3.095204933559535e-05,
+      "loss": 1.0935,
+      "step": 10067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5542073863166346,
+      "learning_rate": 3.094799862572532e-05,
+      "loss": 0.7104,
+      "step": 10068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7513855131503977,
+      "learning_rate": 3.09439477503249e-05,
+      "loss": 1.0381,
+      "step": 10069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.739947473420064,
+      "learning_rate": 3.093989670950684e-05,
+      "loss": 1.3125,
+      "step": 10070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1928497293026568,
+      "learning_rate": 3.093584550338388e-05,
+      "loss": 0.693,
+      "step": 10071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.798442770763447,
+      "learning_rate": 3.093179413206875e-05,
+      "loss": 0.793,
+      "step": 10072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4289273854658493,
+      "learning_rate": 3.092774259567421e-05,
+      "loss": 1.0713,
+      "step": 10073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.028941121670761,
+      "learning_rate": 3.0923690894312994e-05,
+      "loss": 1.0977,
+      "step": 10074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3656417237046057,
+      "learning_rate": 3.0919639028097887e-05,
+      "loss": 1.2842,
+      "step": 10075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.548519925019652,
+      "learning_rate": 3.091558699714163e-05,
+      "loss": 1.2617,
+      "step": 10076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.45031910933247,
+      "learning_rate": 3.0911534801557e-05,
+      "loss": 1.5605,
+      "step": 10077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.996419323910266,
+      "learning_rate": 3.0907482441456746e-05,
+      "loss": 0.3573,
+      "step": 10078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.450709065726959,
+      "learning_rate": 3.090342991695367e-05,
+      "loss": 0.8789,
+      "step": 10079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4279433488097073,
+      "learning_rate": 3.089937722816055e-05,
+      "loss": 0.7031,
+      "step": 10080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.07097139156013,
+      "learning_rate": 3.089532437519016e-05,
+      "loss": 0.9941,
+      "step": 10081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.3242811797219725,
+      "learning_rate": 3.0891271358155286e-05,
+      "loss": 1.6426,
+      "step": 10082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.113798435956969,
+      "learning_rate": 3.088721817716874e-05,
+      "loss": 0.7031,
+      "step": 10083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.573101073129045,
+      "learning_rate": 3.08831648323433e-05,
+      "loss": 0.8262,
+      "step": 10084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.76911373326127,
+      "learning_rate": 3.087911132379178e-05,
+      "loss": 1.2207,
+      "step": 10085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.690124206369851,
+      "learning_rate": 3.0875057651626974e-05,
+      "loss": 1.1084,
+      "step": 10086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.365569495665413,
+      "learning_rate": 3.0871003815961723e-05,
+      "loss": 0.7859,
+      "step": 10087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.724922102068481,
+      "learning_rate": 3.086694981690882e-05,
+      "loss": 0.9346,
+      "step": 10088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.554504574634027,
+      "learning_rate": 3.0862895654581084e-05,
+      "loss": 1.0898,
+      "step": 10089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2183802123748912,
+      "learning_rate": 3.085884132909136e-05,
+      "loss": 0.551,
+      "step": 10090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.907330224345982,
+      "learning_rate": 3.0854786840552464e-05,
+      "loss": 0.9541,
+      "step": 10091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8285887954365925,
+      "learning_rate": 3.0850732189077236e-05,
+      "loss": 0.824,
+      "step": 10092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.869819285588135,
+      "learning_rate": 3.08466773747785e-05,
+      "loss": 0.8174,
+      "step": 10093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.021722250685597,
+      "learning_rate": 3.084262239776913e-05,
+      "loss": 0.8516,
+      "step": 10094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.657653527548588,
+      "learning_rate": 3.0838567258161944e-05,
+      "loss": 1.0459,
+      "step": 10095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.209577226729433,
+      "learning_rate": 3.083451195606982e-05,
+      "loss": 1.1406,
+      "step": 10096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.350068157128996,
+      "learning_rate": 3.0830456491605585e-05,
+      "loss": 1.3105,
+      "step": 10097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4180299815866233,
+      "learning_rate": 3.082640086488214e-05,
+      "loss": 0.8931,
+      "step": 10098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.514366460190635,
+      "learning_rate": 3.082234507601232e-05,
+      "loss": 1.4141,
+      "step": 10099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.337288861971973,
+      "learning_rate": 3.0818289125109015e-05,
+      "loss": 1.2188,
+      "step": 10100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3176119209267387,
+      "learning_rate": 3.081423301228508e-05,
+      "loss": 1.0303,
+      "step": 10101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7380012227537045,
+      "learning_rate": 3.0810176737653414e-05,
+      "loss": 1.4707,
+      "step": 10102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.808638449956476,
+      "learning_rate": 3.080612030132689e-05,
+      "loss": 1.5127,
+      "step": 10103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1002476773207746,
+      "learning_rate": 3.080206370341841e-05,
+      "loss": 0.2216,
+      "step": 10104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.975954602540398,
+      "learning_rate": 3.079800694404086e-05,
+      "loss": 0.665,
+      "step": 10105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3404934447615795,
+      "learning_rate": 3.079395002330713e-05,
+      "loss": 1.1426,
+      "step": 10106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0007170345513003,
+      "learning_rate": 3.078989294133013e-05,
+      "loss": 0.5813,
+      "step": 10107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.252174914548487,
+      "learning_rate": 3.0785835698222765e-05,
+      "loss": 1.6934,
+      "step": 10108
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.098006270798986,
+      "learning_rate": 3.0781778294097954e-05,
+      "loss": 0.6029,
+      "step": 10109
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2706362320873894,
+      "learning_rate": 3.07777207290686e-05,
+      "loss": 0.8569,
+      "step": 10110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.50518022279966,
+      "learning_rate": 3.077366300324764e-05,
+      "loss": 1.4932,
+      "step": 10111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.211850897621842,
+      "learning_rate": 3.076960511674798e-05,
+      "loss": 1.0469,
+      "step": 10112
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3982915769394335,
+      "learning_rate": 3.0765547069682565e-05,
+      "loss": 1.373,
+      "step": 10113
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.455152832903307,
+      "learning_rate": 3.076148886216432e-05,
+      "loss": 1.0996,
+      "step": 10114
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4371995985325365,
+      "learning_rate": 3.075743049430618e-05,
+      "loss": 1.0391,
+      "step": 10115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3871037099399928,
+      "learning_rate": 3.07533719662211e-05,
+      "loss": 1.1416,
+      "step": 10116
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.689373573888302,
+      "learning_rate": 3.074931327802202e-05,
+      "loss": 1.6597,
+      "step": 10117
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.349401803262835,
+      "learning_rate": 3.074525442982189e-05,
+      "loss": 0.7256,
+      "step": 10118
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6280586043780785,
+      "learning_rate": 3.074119542173366e-05,
+      "loss": 1.0049,
+      "step": 10119
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5854691833359253,
+      "learning_rate": 3.0737136253870315e-05,
+      "loss": 1.2988,
+      "step": 10120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.228600199428464,
+      "learning_rate": 3.073307692634479e-05,
+      "loss": 1.6582,
+      "step": 10121
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.174386819816345,
+      "learning_rate": 3.072901743927008e-05,
+      "loss": 1.334,
+      "step": 10122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.276696672658892,
+      "learning_rate": 3.072495779275913e-05,
+      "loss": 1.666,
+      "step": 10123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0072494970751413,
+      "learning_rate": 3.0720897986924956e-05,
+      "loss": 1.2598,
+      "step": 10124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8836676923748317,
+      "learning_rate": 3.07168380218805e-05,
+      "loss": 0.9824,
+      "step": 10125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.410509278395686,
+      "learning_rate": 3.0712777897738785e-05,
+      "loss": 1.4473,
+      "step": 10126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.010935451266439,
+      "learning_rate": 3.070871761461278e-05,
+      "loss": 0.7842,
+      "step": 10127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9995943492541717,
+      "learning_rate": 3.070465717261549e-05,
+      "loss": 1.2207,
+      "step": 10128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9241412687325727,
+      "learning_rate": 3.0700596571859906e-05,
+      "loss": 0.8772,
+      "step": 10129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9483118723367525,
+      "learning_rate": 3.069653581245905e-05,
+      "loss": 0.7812,
+      "step": 10130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.908639119612665,
+      "learning_rate": 3.0692474894525915e-05,
+      "loss": 1.1372,
+      "step": 10131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.86107603438976,
+      "learning_rate": 3.068841381817352e-05,
+      "loss": 1.1279,
+      "step": 10132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.261091958189792,
+      "learning_rate": 3.0684352583514895e-05,
+      "loss": 0.9258,
+      "step": 10133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.17091425249829,
+      "learning_rate": 3.0680291190663044e-05,
+      "loss": 1.8174,
+      "step": 10134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.190164959064179,
+      "learning_rate": 3.0676229639730994e-05,
+      "loss": 1.1074,
+      "step": 10135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9576928067662434,
+      "learning_rate": 3.06721679308318e-05,
+      "loss": 1.2305,
+      "step": 10136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.447152175898208,
+      "learning_rate": 3.0668106064078464e-05,
+      "loss": 1.1602,
+      "step": 10137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.474367425317205,
+      "learning_rate": 3.066404403958406e-05,
+      "loss": 1.6094,
+      "step": 10138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.177497504365873,
+      "learning_rate": 3.0659981857461606e-05,
+      "loss": 1.2217,
+      "step": 10139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5236553653169573,
+      "learning_rate": 3.0655919517824164e-05,
+      "loss": 1.1816,
+      "step": 10140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.749104961046964,
+      "learning_rate": 3.065185702078479e-05,
+      "loss": 0.8872,
+      "step": 10141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8543317342711627,
+      "learning_rate": 3.064779436645653e-05,
+      "loss": 0.8396,
+      "step": 10142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.64157786199213,
+      "learning_rate": 3.064373155495246e-05,
+      "loss": 0.7148,
+      "step": 10143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.39447628785821,
+      "learning_rate": 3.063966858638562e-05,
+      "loss": 1.0928,
+      "step": 10144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.161251524309822,
+      "learning_rate": 3.063560546086911e-05,
+      "loss": 1.4072,
+      "step": 10145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.226137738290307,
+      "learning_rate": 3.063154217851599e-05,
+      "loss": 1.2822,
+      "step": 10146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.273416523649975,
+      "learning_rate": 3.062747873943936e-05,
+      "loss": 1.3691,
+      "step": 10147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.606090659977612,
+      "learning_rate": 3.062341514375227e-05,
+      "loss": 1.292,
+      "step": 10148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.849609159470864,
+      "learning_rate": 3.061935139156783e-05,
+      "loss": 1.9277,
+      "step": 10149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.250701053333345,
+      "learning_rate": 3.061528748299913e-05,
+      "loss": 1.0444,
+      "step": 10150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.18161461523495,
+      "learning_rate": 3.061122341815927e-05,
+      "loss": 1.2686,
+      "step": 10151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.526112693322944,
+      "learning_rate": 3.060715919716134e-05,
+      "loss": 1.2041,
+      "step": 10152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.814567608358163,
+      "learning_rate": 3.060309482011845e-05,
+      "loss": 1.3281,
+      "step": 10153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8631637184294227,
+      "learning_rate": 3.059903028714372e-05,
+      "loss": 0.623,
+      "step": 10154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2756267899839115,
+      "learning_rate": 3.059496559835026e-05,
+      "loss": 1.0521,
+      "step": 10155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.6225799722501835,
+      "learning_rate": 3.059090075385117e-05,
+      "loss": 1.1201,
+      "step": 10156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9804893545529025,
+      "learning_rate": 3.05868357537596e-05,
+      "loss": 1.6328,
+      "step": 10157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6392508096855405,
+      "learning_rate": 3.058277059818866e-05,
+      "loss": 1.5059,
+      "step": 10158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.658460258475258,
+      "learning_rate": 3.0578705287251484e-05,
+      "loss": 0.644,
+      "step": 10159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.6987903582998705,
+      "learning_rate": 3.057463982106123e-05,
+      "loss": 0.915,
+      "step": 10160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.614319919390229,
+      "learning_rate": 3.0570574199730996e-05,
+      "loss": 1.3105,
+      "step": 10161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.028426211992375,
+      "learning_rate": 3.056650842337396e-05,
+      "loss": 1.4766,
+      "step": 10162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.073030580492891,
+      "learning_rate": 3.056244249210326e-05,
+      "loss": 0.7573,
+      "step": 10163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5939061680854585,
+      "learning_rate": 3.055837640603205e-05,
+      "loss": 1.1377,
+      "step": 10164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.948393371181127,
+      "learning_rate": 3.055431016527348e-05,
+      "loss": 1.5566,
+      "step": 10165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.256510009928295,
+      "learning_rate": 3.055024376994073e-05,
+      "loss": 0.9551,
+      "step": 10166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.946003405817479,
+      "learning_rate": 3.0546177220146946e-05,
+      "loss": 0.7559,
+      "step": 10167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1116934498467534,
+      "learning_rate": 3.054211051600532e-05,
+      "loss": 0.8789,
+      "step": 10168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.010301387563997,
+      "learning_rate": 3.0538043657629e-05,
+      "loss": 0.8872,
+      "step": 10169
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.794969133128157,
+      "learning_rate": 3.053397664513119e-05,
+      "loss": 0.8628,
+      "step": 10170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5098728169931293,
+      "learning_rate": 3.0529909478625066e-05,
+      "loss": 1.0518,
+      "step": 10171
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.244908815058746,
+      "learning_rate": 3.05258421582238e-05,
+      "loss": 1.3379,
+      "step": 10172
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.669332443394673,
+      "learning_rate": 3.0521774684040605e-05,
+      "loss": 0.9062,
+      "step": 10173
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.02540286730871,
+      "learning_rate": 3.051770705618866e-05,
+      "loss": 0.7832,
+      "step": 10174
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.107677992140803,
+      "learning_rate": 3.051363927478118e-05,
+      "loss": 0.8486,
+      "step": 10175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9495862717540773,
+      "learning_rate": 3.0509571339931365e-05,
+      "loss": 1.1846,
+      "step": 10176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.143583793371087,
+      "learning_rate": 3.0505503251752425e-05,
+      "loss": 1.2461,
+      "step": 10177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.811030048188296,
+      "learning_rate": 3.0501435010357564e-05,
+      "loss": 0.9409,
+      "step": 10178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3254033507901304,
+      "learning_rate": 3.049736661586001e-05,
+      "loss": 0.5874,
+      "step": 10179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.791577228612765,
+      "learning_rate": 3.0493298068372983e-05,
+      "loss": 1.0352,
+      "step": 10180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.707840187556007,
+      "learning_rate": 3.0489229368009702e-05,
+      "loss": 1.0469,
+      "step": 10181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7873435827597666,
+      "learning_rate": 3.0485160514883405e-05,
+      "loss": 0.8408,
+      "step": 10182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.278132549483134,
+      "learning_rate": 3.0481091509107324e-05,
+      "loss": 0.8779,
+      "step": 10183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9194394938577914,
+      "learning_rate": 3.0477022350794697e-05,
+      "loss": 1.3184,
+      "step": 10184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.089587750209027,
+      "learning_rate": 3.0472953040058772e-05,
+      "loss": 2.332,
+      "step": 10185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0161668359869163,
+      "learning_rate": 3.046888357701278e-05,
+      "loss": 1.384,
+      "step": 10186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6311914904108,
+      "learning_rate": 3.0464813961769993e-05,
+      "loss": 0.8936,
+      "step": 10187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.269084044940281,
+      "learning_rate": 3.046074419444366e-05,
+      "loss": 1.0522,
+      "step": 10188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.589390254468554,
+      "learning_rate": 3.045667427514704e-05,
+      "loss": 0.5103,
+      "step": 10189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4486728771735757,
+      "learning_rate": 3.045260420399339e-05,
+      "loss": 0.7998,
+      "step": 10190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7243024902659916,
+      "learning_rate": 3.0448533981095988e-05,
+      "loss": 0.9612,
+      "step": 10191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0820091489724604,
+      "learning_rate": 3.044446360656811e-05,
+      "loss": 0.8486,
+      "step": 10192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.095931667298721,
+      "learning_rate": 3.0440393080523016e-05,
+      "loss": 0.7744,
+      "step": 10193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2282560633651016,
+      "learning_rate": 3.0436322403074003e-05,
+      "loss": 0.725,
+      "step": 10194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.941611941088912,
+      "learning_rate": 3.043225157433434e-05,
+      "loss": 0.7448,
+      "step": 10195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7169111325160684,
+      "learning_rate": 3.0428180594417337e-05,
+      "loss": 1.0728,
+      "step": 10196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1185468413140063,
+      "learning_rate": 3.042410946343628e-05,
+      "loss": 0.7363,
+      "step": 10197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.843941818307465,
+      "learning_rate": 3.0420038181504464e-05,
+      "loss": 1.2188,
+      "step": 10198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.132969312435283,
+      "learning_rate": 3.0415966748735182e-05,
+      "loss": 1.832,
+      "step": 10199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.022587070924922,
+      "learning_rate": 3.041189516524176e-05,
+      "loss": 1.0658,
+      "step": 10200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1060075016655757,
+      "learning_rate": 3.0407823431137493e-05,
+      "loss": 1.0029,
+      "step": 10201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.766614661436516,
+      "learning_rate": 3.0403751546535708e-05,
+      "loss": 0.6626,
+      "step": 10202
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9494182594976266,
+      "learning_rate": 3.0399679511549718e-05,
+      "loss": 0.6616,
+      "step": 10203
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.827377389126399,
+      "learning_rate": 3.0395607326292836e-05,
+      "loss": 0.9099,
+      "step": 10204
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.066758960447842,
+      "learning_rate": 3.03915349908784e-05,
+      "loss": 1.7051,
+      "step": 10205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.055790921933896,
+      "learning_rate": 3.038746250541975e-05,
+      "loss": 0.6871,
+      "step": 10206
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.136858612545273,
+      "learning_rate": 3.0383389870030205e-05,
+      "loss": 0.8989,
+      "step": 10207
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.928212517184979,
+      "learning_rate": 3.0379317084823118e-05,
+      "loss": 0.4246,
+      "step": 10208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.531117282731895,
+      "learning_rate": 3.0375244149911815e-05,
+      "loss": 1.1797,
+      "step": 10209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.3214609731809,
+      "learning_rate": 3.037117106540966e-05,
+      "loss": 0.8652,
+      "step": 10210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4314555024913993,
+      "learning_rate": 3.036709783143001e-05,
+      "loss": 1.2227,
+      "step": 10211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.741990317492462,
+      "learning_rate": 3.03630244480862e-05,
+      "loss": 1.4346,
+      "step": 10212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9006496774312671,
+      "learning_rate": 3.0358950915491602e-05,
+      "loss": 0.7842,
+      "step": 10213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.895008986311896,
+      "learning_rate": 3.035487723375959e-05,
+      "loss": 1.1953,
+      "step": 10214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3429577325584545,
+      "learning_rate": 3.0350803403003523e-05,
+      "loss": 1.5762,
+      "step": 10215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.153924506554295,
+      "learning_rate": 3.0346729423336772e-05,
+      "loss": 1.1621,
+      "step": 10216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.084941268935331,
+      "learning_rate": 3.0342655294872725e-05,
+      "loss": 0.4507,
+      "step": 10217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.257639639791718,
+      "learning_rate": 3.0338581017724743e-05,
+      "loss": 1.7148,
+      "step": 10218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.4923552652456555,
+      "learning_rate": 3.0334506592006233e-05,
+      "loss": 1.3379,
+      "step": 10219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.178993491052767,
+      "learning_rate": 3.033043201783058e-05,
+      "loss": 1.9766,
+      "step": 10220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4591549184741126,
+      "learning_rate": 3.032635729531117e-05,
+      "loss": 1.2148,
+      "step": 10221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9423140124779612,
+      "learning_rate": 3.0322282424561404e-05,
+      "loss": 0.5526,
+      "step": 10222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.44495083060715,
+      "learning_rate": 3.0318207405694683e-05,
+      "loss": 1.2705,
+      "step": 10223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.10953911036954,
+      "learning_rate": 3.0314132238824415e-05,
+      "loss": 0.6425,
+      "step": 10224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.005036418014962,
+      "learning_rate": 3.0310056924064017e-05,
+      "loss": 1.1904,
+      "step": 10225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.695153054372082,
+      "learning_rate": 3.030598146152689e-05,
+      "loss": 1.5176,
+      "step": 10226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.815172663351433,
+      "learning_rate": 3.030190585132646e-05,
+      "loss": 0.959,
+      "step": 10227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.54968001671843,
+      "learning_rate": 3.0297830093576153e-05,
+      "loss": 1.375,
+      "step": 10228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7343709019221523,
+      "learning_rate": 3.029375418838939e-05,
+      "loss": 0.8423,
+      "step": 10229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.170186765693668,
+      "learning_rate": 3.02896781358796e-05,
+      "loss": 1.2246,
+      "step": 10230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.349983775306047,
+      "learning_rate": 3.0285601936160223e-05,
+      "loss": 0.5259,
+      "step": 10231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0896230587470703,
+      "learning_rate": 3.0281525589344694e-05,
+      "loss": 0.9795,
+      "step": 10232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.990371608001721,
+      "learning_rate": 3.0277449095546455e-05,
+      "loss": 1.5723,
+      "step": 10233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.935765443657753,
+      "learning_rate": 3.0273372454878963e-05,
+      "loss": 0.4971,
+      "step": 10234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8931078973805713,
+      "learning_rate": 3.0269295667455662e-05,
+      "loss": 1.3691,
+      "step": 10235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3190700916720672,
+      "learning_rate": 3.0265218733390005e-05,
+      "loss": 0.7225,
+      "step": 10236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7155443783152147,
+      "learning_rate": 3.026114165279545e-05,
+      "loss": 1.2031,
+      "step": 10237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.434658316476364,
+      "learning_rate": 3.025706442578547e-05,
+      "loss": 1.0596,
+      "step": 10238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0856689966103517,
+      "learning_rate": 3.0252987052473526e-05,
+      "loss": 1.3281,
+      "step": 10239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2239820307947897,
+      "learning_rate": 3.0248909532973092e-05,
+      "loss": 1.0234,
+      "step": 10240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1777425859920876,
+      "learning_rate": 3.024483186739764e-05,
+      "loss": 1.1418,
+      "step": 10241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7478813333001506,
+      "learning_rate": 3.0240754055860655e-05,
+      "loss": 1.4102,
+      "step": 10242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4824181293306222,
+      "learning_rate": 3.0236676098475612e-05,
+      "loss": 1.4155,
+      "step": 10243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6834668878936574,
+      "learning_rate": 3.0232597995356017e-05,
+      "loss": 1.3535,
+      "step": 10244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.640652948674256,
+      "learning_rate": 3.0228519746615337e-05,
+      "loss": 0.9844,
+      "step": 10245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8448678846534137,
+      "learning_rate": 3.0224441352367082e-05,
+      "loss": 0.7578,
+      "step": 10246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.518315079389553,
+      "learning_rate": 3.022036281272476e-05,
+      "loss": 1.1035,
+      "step": 10247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.0003278121136665,
+      "learning_rate": 3.0216284127801863e-05,
+      "loss": 0.7764,
+      "step": 10248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.927837767142479,
+      "learning_rate": 3.0212205297711903e-05,
+      "loss": 0.8594,
+      "step": 10249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7991783843988127,
+      "learning_rate": 3.0208126322568385e-05,
+      "loss": 0.9658,
+      "step": 10250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8000525639913443,
+      "learning_rate": 3.020404720248484e-05,
+      "loss": 1.1826,
+      "step": 10251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.077768595237915,
+      "learning_rate": 3.0199967937574774e-05,
+      "loss": 1.207,
+      "step": 10252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.733198668126874,
+      "learning_rate": 3.019588852795172e-05,
+      "loss": 0.8423,
+      "step": 10253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9639727442223371,
+      "learning_rate": 3.01918089737292e-05,
+      "loss": 0.8137,
+      "step": 10254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.375440712537336,
+      "learning_rate": 3.0187729275020754e-05,
+      "loss": 0.9556,
+      "step": 10255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.259522708786166,
+      "learning_rate": 3.0183649431939916e-05,
+      "loss": 0.6299,
+      "step": 10256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.295095679379569,
+      "learning_rate": 3.0179569444600218e-05,
+      "loss": 0.875,
+      "step": 10257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9409667316151817,
+      "learning_rate": 3.017548931311522e-05,
+      "loss": 1.168,
+      "step": 10258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5793469603798904,
+      "learning_rate": 3.0171409037598453e-05,
+      "loss": 0.5439,
+      "step": 10259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.924119419437831,
+      "learning_rate": 3.016732861816348e-05,
+      "loss": 1.3047,
+      "step": 10260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.729082227570303,
+      "learning_rate": 3.0163248054923853e-05,
+      "loss": 1.1211,
+      "step": 10261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.864445247736589,
+      "learning_rate": 3.0159167347993146e-05,
+      "loss": 1.1426,
+      "step": 10262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.8785923168865795,
+      "learning_rate": 3.0155086497484897e-05,
+      "loss": 1.1602,
+      "step": 10263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1364028410540747,
+      "learning_rate": 3.0151005503512698e-05,
+      "loss": 0.5376,
+      "step": 10264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.114507028459834,
+      "learning_rate": 3.0146924366190104e-05,
+      "loss": 0.9888,
+      "step": 10265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.93196552599893,
+      "learning_rate": 3.014284308563071e-05,
+      "loss": 0.873,
+      "step": 10266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.51508946375296,
+      "learning_rate": 3.0138761661948083e-05,
+      "loss": 1.0371,
+      "step": 10267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5766186561734004,
+      "learning_rate": 3.0134680095255817e-05,
+      "loss": 1.5527,
+      "step": 10268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.457544536271811,
+      "learning_rate": 3.013059838566748e-05,
+      "loss": 1.0957,
+      "step": 10269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4621418261609844,
+      "learning_rate": 3.0126516533296695e-05,
+      "loss": 0.783,
+      "step": 10270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4364272471305815,
+      "learning_rate": 3.0122434538257028e-05,
+      "loss": 0.8164,
+      "step": 10271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.272148560477284,
+      "learning_rate": 3.0118352400662103e-05,
+      "loss": 1.2148,
+      "step": 10272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2417284133329547,
+      "learning_rate": 3.011427012062551e-05,
+      "loss": 0.9966,
+      "step": 10273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.767620131804904,
+      "learning_rate": 3.0110187698260863e-05,
+      "loss": 1.3135,
+      "step": 10274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.111162459713826,
+      "learning_rate": 3.0106105133681765e-05,
+      "loss": 1.4297,
+      "step": 10275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.127226504577292,
+      "learning_rate": 3.0102022427001853e-05,
+      "loss": 0.5957,
+      "step": 10276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.468253572699588,
+      "learning_rate": 3.009793957833473e-05,
+      "loss": 0.7076,
+      "step": 10277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.538361097394054,
+      "learning_rate": 3.0093856587794017e-05,
+      "loss": 1.9668,
+      "step": 10278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.441334993917369,
+      "learning_rate": 3.0089773455493357e-05,
+      "loss": 0.8772,
+      "step": 10279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8969050372037783,
+      "learning_rate": 3.0085690181546366e-05,
+      "loss": 0.4253,
+      "step": 10280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.777439568744728,
+      "learning_rate": 3.0081606766066694e-05,
+      "loss": 1.2793,
+      "step": 10281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.652297857450184,
+      "learning_rate": 3.0077523209167974e-05,
+      "loss": 2.1016,
+      "step": 10282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0383202595167034,
+      "learning_rate": 3.0073439510963846e-05,
+      "loss": 0.9165,
+      "step": 10283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9604129376925976,
+      "learning_rate": 3.0069355671567962e-05,
+      "loss": 0.9785,
+      "step": 10284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8654721093588877,
+      "learning_rate": 3.0065271691093977e-05,
+      "loss": 0.8508,
+      "step": 10285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.188604398158135,
+      "learning_rate": 3.006118756965554e-05,
+      "loss": 0.9199,
+      "step": 10286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.970369384069884,
+      "learning_rate": 3.0057103307366317e-05,
+      "loss": 1.4863,
+      "step": 10287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.263865662542084,
+      "learning_rate": 3.0053018904339962e-05,
+      "loss": 1.0879,
+      "step": 10288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3903190067904396,
+      "learning_rate": 3.0048934360690155e-05,
+      "loss": 1.2695,
+      "step": 10289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.005515955988718,
+      "learning_rate": 3.0044849676530556e-05,
+      "loss": 1.3467,
+      "step": 10290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8027881364617313,
+      "learning_rate": 3.0040764851974846e-05,
+      "loss": 1.7969,
+      "step": 10291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.447897750969122,
+      "learning_rate": 3.0036679887136697e-05,
+      "loss": 1.2168,
+      "step": 10292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.104292352849984,
+      "learning_rate": 3.00325947821298e-05,
+      "loss": 1.1885,
+      "step": 10293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1760402447146925,
+      "learning_rate": 3.0028509537067844e-05,
+      "loss": 0.7627,
+      "step": 10294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.916797403673146,
+      "learning_rate": 3.002442415206451e-05,
+      "loss": 1.7891,
+      "step": 10295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1516796825359004,
+      "learning_rate": 3.0020338627233503e-05,
+      "loss": 0.957,
+      "step": 10296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4777839608773413,
+      "learning_rate": 3.0016252962688508e-05,
+      "loss": 0.7417,
+      "step": 10297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.981857846105641,
+      "learning_rate": 3.0012167158543243e-05,
+      "loss": 1.0732,
+      "step": 10298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2208378510897897,
+      "learning_rate": 3.0008081214911398e-05,
+      "loss": 1.6094,
+      "step": 10299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0653346430690114,
+      "learning_rate": 3.00039951319067e-05,
+      "loss": 1.0186,
+      "step": 10300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.348459082964229,
+      "learning_rate": 2.9999908909642844e-05,
+      "loss": 0.6182,
+      "step": 10301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1358892026518226,
+      "learning_rate": 2.9995822548233566e-05,
+      "loss": 0.7324,
+      "step": 10302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6191579205128983,
+      "learning_rate": 2.9991736047792575e-05,
+      "loss": 1.2773,
+      "step": 10303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5590517988131443,
+      "learning_rate": 2.998764940843361e-05,
+      "loss": 0.9297,
+      "step": 10304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0326505087742732,
+      "learning_rate": 2.998356263027039e-05,
+      "loss": 1.2188,
+      "step": 10305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.324036285018726,
+      "learning_rate": 2.9979475713416642e-05,
+      "loss": 1.1865,
+      "step": 10306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7925062060813994,
+      "learning_rate": 2.997538865798612e-05,
+      "loss": 1.0918,
+      "step": 10307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9393314811306044,
+      "learning_rate": 2.9971301464092562e-05,
+      "loss": 0.353,
+      "step": 10308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.668285407015881,
+      "learning_rate": 2.9967214131849696e-05,
+      "loss": 1.3926,
+      "step": 10309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.994262613199076,
+      "learning_rate": 2.9963126661371293e-05,
+      "loss": 1.7109,
+      "step": 10310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.270933990532897,
+      "learning_rate": 2.9959039052771087e-05,
+      "loss": 1.3535,
+      "step": 10311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.844671743640179,
+      "learning_rate": 2.9954951306162847e-05,
+      "loss": 1.1357,
+      "step": 10312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.575653554793152,
+      "learning_rate": 2.995086342166033e-05,
+      "loss": 1.1133,
+      "step": 10313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.247820010035472,
+      "learning_rate": 2.9946775399377303e-05,
+      "loss": 0.8555,
+      "step": 10314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9649386389892225,
+      "learning_rate": 2.9942687239427525e-05,
+      "loss": 1.1562,
+      "step": 10315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8167203815311797,
+      "learning_rate": 2.9938598941924774e-05,
+      "loss": 0.9922,
+      "step": 10316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.353710828101659,
+      "learning_rate": 2.9934510506982826e-05,
+      "loss": 0.4935,
+      "step": 10317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.435807018972463,
+      "learning_rate": 2.9930421934715458e-05,
+      "loss": 0.8909,
+      "step": 10318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.182820573926997,
+      "learning_rate": 2.9926333225236464e-05,
+      "loss": 0.9121,
+      "step": 10319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.19817190511258,
+      "learning_rate": 2.9922244378659604e-05,
+      "loss": 0.835,
+      "step": 10320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.01697961990432,
+      "learning_rate": 2.9918155395098702e-05,
+      "loss": 0.9717,
+      "step": 10321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3701188975902876,
+      "learning_rate": 2.991406627466753e-05,
+      "loss": 1.0381,
+      "step": 10322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.772488610284075,
+      "learning_rate": 2.99099770174799e-05,
+      "loss": 1.6699,
+      "step": 10323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.043493841607671,
+      "learning_rate": 2.9905887623649602e-05,
+      "loss": 1.1768,
+      "step": 10324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.385737090821664,
+      "learning_rate": 2.990179809329045e-05,
+      "loss": 1.4258,
+      "step": 10325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.027345649926786,
+      "learning_rate": 2.9897708426516253e-05,
+      "loss": 1.6816,
+      "step": 10326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.701454660834769,
+      "learning_rate": 2.9893618623440827e-05,
+      "loss": 0.8215,
+      "step": 10327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1599202639379596,
+      "learning_rate": 2.9889528684177987e-05,
+      "loss": 0.219,
+      "step": 10328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.878569629930964,
+      "learning_rate": 2.9885438608841547e-05,
+      "loss": 1.8457,
+      "step": 10329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.467271289526019,
+      "learning_rate": 2.988134839754535e-05,
+      "loss": 0.6299,
+      "step": 10330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6080545759704292,
+      "learning_rate": 2.9877258050403212e-05,
+      "loss": 0.7217,
+      "step": 10331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0273986691366215,
+      "learning_rate": 2.9873167567528958e-05,
+      "loss": 1.1909,
+      "step": 10332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6418215377546996,
+      "learning_rate": 2.9869076949036445e-05,
+      "loss": 1.1484,
+      "step": 10333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0862413963657014,
+      "learning_rate": 2.9864986195039502e-05,
+      "loss": 1.377,
+      "step": 10334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.602003921553538,
+      "learning_rate": 2.986089530565197e-05,
+      "loss": 0.7148,
+      "step": 10335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4630729008865706,
+      "learning_rate": 2.9856804280987706e-05,
+      "loss": 1.3379,
+      "step": 10336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.164468169403196,
+      "learning_rate": 2.9852713121160547e-05,
+      "loss": 1.0029,
+      "step": 10337
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9498943439248504,
+      "learning_rate": 2.9848621826284362e-05,
+      "loss": 0.8877,
+      "step": 10338
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2104037321516254,
+      "learning_rate": 2.9844530396473003e-05,
+      "loss": 0.5063,
+      "step": 10339
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.832897407779931,
+      "learning_rate": 2.9840438831840338e-05,
+      "loss": 1.2168,
+      "step": 10340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1815280695349473,
+      "learning_rate": 2.9836347132500232e-05,
+      "loss": 1.7598,
+      "step": 10341
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.905010984731578,
+      "learning_rate": 2.9832255298566557e-05,
+      "loss": 1.2832,
+      "step": 10342
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.197388247012924,
+      "learning_rate": 2.9828163330153174e-05,
+      "loss": 1.375,
+      "step": 10343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.632693475955121,
+      "learning_rate": 2.982407122737398e-05,
+      "loss": 0.7861,
+      "step": 10344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.239486290063815,
+      "learning_rate": 2.9819978990342844e-05,
+      "loss": 1.0938,
+      "step": 10345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.13292807555039,
+      "learning_rate": 2.981588661917366e-05,
+      "loss": 1.9355,
+      "step": 10346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.557952923656805,
+      "learning_rate": 2.9811794113980302e-05,
+      "loss": 1.4434,
+      "step": 10347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.936938292995826,
+      "learning_rate": 2.980770147487668e-05,
+      "loss": 1.1191,
+      "step": 10348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8780073208366599,
+      "learning_rate": 2.980360870197668e-05,
+      "loss": 1.0964,
+      "step": 10349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.724684541273868,
+      "learning_rate": 2.9799515795394206e-05,
+      "loss": 1.0996,
+      "step": 10350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0001803979676156,
+      "learning_rate": 2.9795422755243168e-05,
+      "loss": 0.9199,
+      "step": 10351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.456012300292886,
+      "learning_rate": 2.9791329581637457e-05,
+      "loss": 0.7358,
+      "step": 10352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.545685421600139,
+      "learning_rate": 2.9787236274691e-05,
+      "loss": 1.1455,
+      "step": 10353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1057056300825914,
+      "learning_rate": 2.97831428345177e-05,
+      "loss": 1.0142,
+      "step": 10354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.11808714157981,
+      "learning_rate": 2.9779049261231495e-05,
+      "loss": 1.3711,
+      "step": 10355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.600083672570965,
+      "learning_rate": 2.9774955554946282e-05,
+      "loss": 1.498,
+      "step": 10356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9519197133732584,
+      "learning_rate": 2.9770861715776007e-05,
+      "loss": 0.8994,
+      "step": 10357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.139154597777396,
+      "learning_rate": 2.9766767743834584e-05,
+      "loss": 0.8828,
+      "step": 10358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.143314760027789,
+      "learning_rate": 2.9762673639235967e-05,
+      "loss": 0.7131,
+      "step": 10359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8638739058966385,
+      "learning_rate": 2.975857940209407e-05,
+      "loss": 1.2021,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.333068923501113,
+      "learning_rate": 2.9754485032522857e-05,
+      "loss": 1.0986,
+      "step": 10361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.706668185373015,
+      "learning_rate": 2.9750390530636246e-05,
+      "loss": 1.5449,
+      "step": 10362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4869981137996175,
+      "learning_rate": 2.9746295896548216e-05,
+      "loss": 0.9102,
+      "step": 10363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.883043068036501,
+      "learning_rate": 2.9742201130372693e-05,
+      "loss": 0.8906,
+      "step": 10364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0789180762446646,
+      "learning_rate": 2.973810623222365e-05,
+      "loss": 1.0713,
+      "step": 10365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9008490547601469,
+      "learning_rate": 2.973401120221504e-05,
+      "loss": 0.832,
+      "step": 10366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.053312035262052,
+      "learning_rate": 2.972991604046082e-05,
+      "loss": 1.2168,
+      "step": 10367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.205327814594035,
+      "learning_rate": 2.9725820747074966e-05,
+      "loss": 0.9897,
+      "step": 10368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.846720320043006,
+      "learning_rate": 2.972172532217144e-05,
+      "loss": 1.0459,
+      "step": 10369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.118209021363951,
+      "learning_rate": 2.971762976586423e-05,
+      "loss": 0.6597,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.285751082771293,
+      "learning_rate": 2.971353407826729e-05,
+      "loss": 1.0059,
+      "step": 10371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6759311160700046,
+      "learning_rate": 2.9709438259494626e-05,
+      "loss": 0.8184,
+      "step": 10372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2522973963787796,
+      "learning_rate": 2.9705342309660205e-05,
+      "loss": 0.8467,
+      "step": 10373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5227503357628946,
+      "learning_rate": 2.9701246228878032e-05,
+      "loss": 1.4551,
+      "step": 10374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.043748987413441,
+      "learning_rate": 2.969715001726208e-05,
+      "loss": 0.6414,
+      "step": 10375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9691408163024047,
+      "learning_rate": 2.9693053674926362e-05,
+      "loss": 0.8789,
+      "step": 10376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6826733561420397,
+      "learning_rate": 2.9688957201984868e-05,
+      "loss": 0.9224,
+      "step": 10377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.190538129823529,
+      "learning_rate": 2.9684860598551612e-05,
+      "loss": 1.0312,
+      "step": 10378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.376900314527056,
+      "learning_rate": 2.9680763864740585e-05,
+      "loss": 1.209,
+      "step": 10379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4588309824202765,
+      "learning_rate": 2.9676667000665805e-05,
+      "loss": 1.4214,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.669988490447765,
+      "learning_rate": 2.9672570006441285e-05,
+      "loss": 0.8262,
+      "step": 10381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.672683084573819,
+      "learning_rate": 2.9668472882181046e-05,
+      "loss": 1.0312,
+      "step": 10382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2933530209309865,
+      "learning_rate": 2.966437562799911e-05,
+      "loss": 0.5059,
+      "step": 10383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.647730640166631,
+      "learning_rate": 2.96602782440095e-05,
+      "loss": 0.6309,
+      "step": 10384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4342528992026713,
+      "learning_rate": 2.9656180730326238e-05,
+      "loss": 0.8535,
+      "step": 10385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5540644734668003,
+      "learning_rate": 2.9652083087063366e-05,
+      "loss": 0.9268,
+      "step": 10386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2601977358035823,
+      "learning_rate": 2.9647985314334913e-05,
+      "loss": 0.9695,
+      "step": 10387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.872811023152002,
+      "learning_rate": 2.9643887412254924e-05,
+      "loss": 1.124,
+      "step": 10388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.0931552134558045,
+      "learning_rate": 2.9639789380937432e-05,
+      "loss": 0.667,
+      "step": 10389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.123830224835475,
+      "learning_rate": 2.9635691220496497e-05,
+      "loss": 1.1152,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.576005839216829,
+      "learning_rate": 2.9631592931046165e-05,
+      "loss": 0.6636,
+      "step": 10391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.991191771803769,
+      "learning_rate": 2.9627494512700482e-05,
+      "loss": 2.1006,
+      "step": 10392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0310425048284393,
+      "learning_rate": 2.9623395965573515e-05,
+      "loss": 1.1846,
+      "step": 10393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4444742372379697,
+      "learning_rate": 2.9619297289779307e-05,
+      "loss": 1.0811,
+      "step": 10394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.044936728530792,
+      "learning_rate": 2.961519848543195e-05,
+      "loss": 0.7754,
+      "step": 10395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.267795864431619,
+      "learning_rate": 2.9611099552645494e-05,
+      "loss": 0.75,
+      "step": 10396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.240179245793446,
+      "learning_rate": 2.9607000491534022e-05,
+      "loss": 1.2891,
+      "step": 10397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9322184705435204,
+      "learning_rate": 2.9602901302211588e-05,
+      "loss": 0.7617,
+      "step": 10398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.315118726212423,
+      "learning_rate": 2.959880198479229e-05,
+      "loss": 0.9951,
+      "step": 10399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.126321017581026,
+      "learning_rate": 2.9594702539390207e-05,
+      "loss": 0.9932,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6241402757540007,
+      "learning_rate": 2.9590602966119423e-05,
+      "loss": 1.2324,
+      "step": 10401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.430457454401286,
+      "learning_rate": 2.9586503265094025e-05,
+      "loss": 1.1738,
+      "step": 10402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6619204378344685,
+      "learning_rate": 2.9582403436428107e-05,
+      "loss": 0.7783,
+      "step": 10403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6741414412508067,
+      "learning_rate": 2.9578303480235774e-05,
+      "loss": 0.6899,
+      "step": 10404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.492365480146499,
+      "learning_rate": 2.9574203396631113e-05,
+      "loss": 1.2344,
+      "step": 10405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.34085482974684,
+      "learning_rate": 2.9570103185728238e-05,
+      "loss": 1.4629,
+      "step": 10406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.557067563314299,
+      "learning_rate": 2.9566002847641246e-05,
+      "loss": 1.1172,
+      "step": 10407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.493493565038716,
+      "learning_rate": 2.956190238248425e-05,
+      "loss": 1.377,
+      "step": 10408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.13441772431819,
+      "learning_rate": 2.9557801790371376e-05,
+      "loss": 0.7344,
+      "step": 10409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6515941483020224,
+      "learning_rate": 2.955370107141674e-05,
+      "loss": 1.2871,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.398220456664809,
+      "learning_rate": 2.9549600225734442e-05,
+      "loss": 1.0977,
+      "step": 10411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1468536709590142,
+      "learning_rate": 2.9545499253438635e-05,
+      "loss": 1.4043,
+      "step": 10412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.283666294969958,
+      "learning_rate": 2.9541398154643418e-05,
+      "loss": 1.3281,
+      "step": 10413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9255957214249295,
+      "learning_rate": 2.9537296929462953e-05,
+      "loss": 1.4961,
+      "step": 10414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.430277780122728,
+      "learning_rate": 2.9533195578011356e-05,
+      "loss": 1.0879,
+      "step": 10415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6913313277780344,
+      "learning_rate": 2.9529094100402778e-05,
+      "loss": 1.4121,
+      "step": 10416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.489680062087351,
+      "learning_rate": 2.9524992496751346e-05,
+      "loss": 1.3867,
+      "step": 10417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8433534115093217,
+      "learning_rate": 2.952089076717122e-05,
+      "loss": 0.9561,
+      "step": 10418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.833590401934221,
+      "learning_rate": 2.951678891177655e-05,
+      "loss": 1.3516,
+      "step": 10419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3641525008462674,
+      "learning_rate": 2.951268693068147e-05,
+      "loss": 1.1221,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.562820821671562,
+      "learning_rate": 2.9508584824000157e-05,
+      "loss": 0.6592,
+      "step": 10421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6242529545751694,
+      "learning_rate": 2.950448259184676e-05,
+      "loss": 0.6589,
+      "step": 10422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.191174412265923,
+      "learning_rate": 2.9500380234335458e-05,
+      "loss": 0.9277,
+      "step": 10423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3583630545353,
+      "learning_rate": 2.9496277751580392e-05,
+      "loss": 1.209,
+      "step": 10424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.029818882193562,
+      "learning_rate": 2.9492175143695762e-05,
+      "loss": 1.1348,
+      "step": 10425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.564699671007469,
+      "learning_rate": 2.948807241079571e-05,
+      "loss": 0.7959,
+      "step": 10426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.378734062490229,
+      "learning_rate": 2.9483969552994438e-05,
+      "loss": 1.1689,
+      "step": 10427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.224558788658825,
+      "learning_rate": 2.9479866570406118e-05,
+      "loss": 1.4141,
+      "step": 10428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8548984192850075,
+      "learning_rate": 2.9475763463144936e-05,
+      "loss": 0.9346,
+      "step": 10429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7727687077163674,
+      "learning_rate": 2.9471660231325076e-05,
+      "loss": 0.7715,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.784788626169749,
+      "learning_rate": 2.946755687506073e-05,
+      "loss": 1.0684,
+      "step": 10431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.575252342822238,
+      "learning_rate": 2.94634533944661e-05,
+      "loss": 0.7948,
+      "step": 10432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.256596692861813,
+      "learning_rate": 2.9459349789655384e-05,
+      "loss": 1.2783,
+      "step": 10433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0030131292422295,
+      "learning_rate": 2.945524606074277e-05,
+      "loss": 1.1582,
+      "step": 10434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.088500705685585,
+      "learning_rate": 2.9451142207842476e-05,
+      "loss": 1.8066,
+      "step": 10435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.366696246951791,
+      "learning_rate": 2.9447038231068703e-05,
+      "loss": 1.4395,
+      "step": 10436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.19888614370416,
+      "learning_rate": 2.9442934130535667e-05,
+      "loss": 1.1787,
+      "step": 10437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.301977394899418,
+      "learning_rate": 2.943882990635759e-05,
+      "loss": 0.7373,
+      "step": 10438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8496145623560563,
+      "learning_rate": 2.9434725558648674e-05,
+      "loss": 0.7373,
+      "step": 10439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.434787708404484,
+      "learning_rate": 2.9430621087523152e-05,
+      "loss": 0.8853,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.036405760286506,
+      "learning_rate": 2.942651649309525e-05,
+      "loss": 1.1689,
+      "step": 10441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.126417436689653,
+      "learning_rate": 2.9422411775479204e-05,
+      "loss": 1.2725,
+      "step": 10442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.01593569405197,
+      "learning_rate": 2.941830693478923e-05,
+      "loss": 1.4004,
+      "step": 10443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3910435328906035,
+      "learning_rate": 2.9414201971139577e-05,
+      "loss": 0.9354,
+      "step": 10444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7977611480537146,
+      "learning_rate": 2.9410096884644473e-05,
+      "loss": 1.106,
+      "step": 10445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7913354326046975,
+      "learning_rate": 2.9405991675418175e-05,
+      "loss": 0.761,
+      "step": 10446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1344371968237805,
+      "learning_rate": 2.940188634357492e-05,
+      "loss": 1.459,
+      "step": 10447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5113547908324327,
+      "learning_rate": 2.9397780889228964e-05,
+      "loss": 1.373,
+      "step": 10448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.181735492226322,
+      "learning_rate": 2.9393675312494552e-05,
+      "loss": 1.0947,
+      "step": 10449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1582784649621467,
+      "learning_rate": 2.9389569613485946e-05,
+      "loss": 1.2109,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.340366908386216,
+      "learning_rate": 2.93854637923174e-05,
+      "loss": 1.3672,
+      "step": 10451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.119986058115147,
+      "learning_rate": 2.9381357849103193e-05,
+      "loss": 1.4785,
+      "step": 10452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.888202817132327,
+      "learning_rate": 2.937725178395757e-05,
+      "loss": 0.707,
+      "step": 10453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1207157522001703,
+      "learning_rate": 2.9373145596994816e-05,
+      "loss": 1.3613,
+      "step": 10454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.503711653225094,
+      "learning_rate": 2.93690392883292e-05,
+      "loss": 0.96,
+      "step": 10455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.984593237838542,
+      "learning_rate": 2.9364932858075e-05,
+      "loss": 1.5918,
+      "step": 10456
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1604011355885806,
+      "learning_rate": 2.93608263063465e-05,
+      "loss": 1.127,
+      "step": 10457
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0893788043326142,
+      "learning_rate": 2.9356719633257974e-05,
+      "loss": 1.4688,
+      "step": 10458
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2736172501550476,
+      "learning_rate": 2.9352612838923708e-05,
+      "loss": 1.2939,
+      "step": 10459
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.2439399598069585,
+      "learning_rate": 2.934850592345801e-05,
+      "loss": 0.8721,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.84253195715219,
+      "learning_rate": 2.934439888697516e-05,
+      "loss": 1.0186,
+      "step": 10461
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.31675892674418,
+      "learning_rate": 2.9340291729589454e-05,
+      "loss": 0.5439,
+      "step": 10462
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2278126202038955,
+      "learning_rate": 2.93361844514152e-05,
+      "loss": 1.04,
+      "step": 10463
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.123364093123804,
+      "learning_rate": 2.9332077052566685e-05,
+      "loss": 1.5488,
+      "step": 10464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.370475474760376,
+      "learning_rate": 2.9327969533158244e-05,
+      "loss": 1.166,
+      "step": 10465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1386018376111604,
+      "learning_rate": 2.9323861893304167e-05,
+      "loss": 1.3086,
+      "step": 10466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.287175813663402,
+      "learning_rate": 2.931975413311878e-05,
+      "loss": 0.6816,
+      "step": 10467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.544738530013019,
+      "learning_rate": 2.9315646252716383e-05,
+      "loss": 1.5508,
+      "step": 10468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.848088640944901,
+      "learning_rate": 2.9311538252211317e-05,
+      "loss": 1.3555,
+      "step": 10469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0958663720420696,
+      "learning_rate": 2.9307430131717894e-05,
+      "loss": 0.6084,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7596379363263974,
+      "learning_rate": 2.9303321891350445e-05,
+      "loss": 0.9014,
+      "step": 10471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.61173137327402,
+      "learning_rate": 2.9299213531223297e-05,
+      "loss": 0.3609,
+      "step": 10472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3260974707712343,
+      "learning_rate": 2.9295105051450784e-05,
+      "loss": 0.8516,
+      "step": 10473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4971129913919725,
+      "learning_rate": 2.9290996452147258e-05,
+      "loss": 0.9853,
+      "step": 10474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8671603659729208,
+      "learning_rate": 2.9286887733427032e-05,
+      "loss": 0.9912,
+      "step": 10475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.08176326998109,
+      "learning_rate": 2.9282778895404478e-05,
+      "loss": 1.1631,
+      "step": 10476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.082807513241183,
+      "learning_rate": 2.927866993819392e-05,
+      "loss": 0.791,
+      "step": 10477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.960705193908805,
+      "learning_rate": 2.927456086190973e-05,
+      "loss": 1.1484,
+      "step": 10478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.940273037365022,
+      "learning_rate": 2.9270451666666242e-05,
+      "loss": 0.8151,
+      "step": 10479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.160802322820976,
+      "learning_rate": 2.926634235257783e-05,
+      "loss": 1.5391,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.497745832806111,
+      "learning_rate": 2.9262232919758847e-05,
+      "loss": 1.2002,
+      "step": 10481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.75563097033651,
+      "learning_rate": 2.9258123368323647e-05,
+      "loss": 0.605,
+      "step": 10482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0172995231926665,
+      "learning_rate": 2.9254013698386617e-05,
+      "loss": 1.252,
+      "step": 10483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6959897604218366,
+      "learning_rate": 2.9249903910062116e-05,
+      "loss": 0.9111,
+      "step": 10484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.845184027118462,
+      "learning_rate": 2.9245794003464515e-05,
+      "loss": 0.5999,
+      "step": 10485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.460954388855399,
+      "learning_rate": 2.9241683978708205e-05,
+      "loss": 1.0049,
+      "step": 10486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.215811627467727,
+      "learning_rate": 2.923757383590755e-05,
+      "loss": 1.6738,
+      "step": 10487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7281090140557107,
+      "learning_rate": 2.923346357517694e-05,
+      "loss": 1.1239,
+      "step": 10488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6935631652486096,
+      "learning_rate": 2.922935319663076e-05,
+      "loss": 1.0698,
+      "step": 10489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.81591718774984,
+      "learning_rate": 2.922524270038341e-05,
+      "loss": 1.1387,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3474082543333727,
+      "learning_rate": 2.922113208654927e-05,
+      "loss": 0.9424,
+      "step": 10491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1935770906736494,
+      "learning_rate": 2.921702135524274e-05,
+      "loss": 1.0171,
+      "step": 10492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.490551456506553,
+      "learning_rate": 2.921291050657823e-05,
+      "loss": 1.1289,
+      "step": 10493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3118005674270234,
+      "learning_rate": 2.920879954067013e-05,
+      "loss": 1.0518,
+      "step": 10494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.655138025978313,
+      "learning_rate": 2.9204688457632856e-05,
+      "loss": 1.2383,
+      "step": 10495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3348253884422383,
+      "learning_rate": 2.92005772575808e-05,
+      "loss": 0.7998,
+      "step": 10496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.543873462174653,
+      "learning_rate": 2.919646594062841e-05,
+      "loss": 0.8691,
+      "step": 10497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8022394156969894,
+      "learning_rate": 2.9192354506890068e-05,
+      "loss": 1.2246,
+      "step": 10498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.555135798419383,
+      "learning_rate": 2.9188242956480217e-05,
+      "loss": 1.0645,
+      "step": 10499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.256455908174021,
+      "learning_rate": 2.918413128951326e-05,
+      "loss": 1.1289,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.240906130215317,
+      "learning_rate": 2.9180019506103633e-05,
+      "loss": 1.0977,
+      "step": 10501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.828034878975661,
+      "learning_rate": 2.9175907606365766e-05,
+      "loss": 1.0605,
+      "step": 10502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.741130456491738,
+      "learning_rate": 2.9171795590414098e-05,
+      "loss": 0.9961,
+      "step": 10503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.192809500516221,
+      "learning_rate": 2.916768345836305e-05,
+      "loss": 0.6108,
+      "step": 10504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.289892573439699,
+      "learning_rate": 2.9163571210327063e-05,
+      "loss": 0.916,
+      "step": 10505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.031910918049188,
+      "learning_rate": 2.9159458846420596e-05,
+      "loss": 0.8489,
+      "step": 10506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.857106961297294,
+      "learning_rate": 2.9155346366758074e-05,
+      "loss": 0.6687,
+      "step": 10507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9921667169377075,
+      "learning_rate": 2.9151233771453956e-05,
+      "loss": 1.0908,
+      "step": 10508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.60925343003183,
+      "learning_rate": 2.9147121060622695e-05,
+      "loss": 2.0273,
+      "step": 10509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4414311522167496,
+      "learning_rate": 2.9143008234378743e-05,
+      "loss": 1.334,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6822441338436702,
+      "learning_rate": 2.9138895292836555e-05,
+      "loss": 1.4277,
+      "step": 10511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8386263120203816,
+      "learning_rate": 2.91347822361106e-05,
+      "loss": 0.6812,
+      "step": 10512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.092766032381614,
+      "learning_rate": 2.9130669064315334e-05,
+      "loss": 0.6675,
+      "step": 10513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.988160134647307,
+      "learning_rate": 2.9126555777565234e-05,
+      "loss": 1.1396,
+      "step": 10514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.802888766299651,
+      "learning_rate": 2.9122442375974756e-05,
+      "loss": 1.3027,
+      "step": 10515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.735269896278415,
+      "learning_rate": 2.91183288596584e-05,
+      "loss": 1.3457,
+      "step": 10516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.404323628113109,
+      "learning_rate": 2.9114215228730616e-05,
+      "loss": 0.9351,
+      "step": 10517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.669916963607635,
+      "learning_rate": 2.9110101483305914e-05,
+      "loss": 0.3506,
+      "step": 10518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.840226390304328,
+      "learning_rate": 2.9105987623498737e-05,
+      "loss": 1.4414,
+      "step": 10519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.9887777920155525,
+      "learning_rate": 2.9101873649423616e-05,
+      "loss": 1.874,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7845482431411925,
+      "learning_rate": 2.909775956119501e-05,
+      "loss": 1.3149,
+      "step": 10521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3889677998227787,
+      "learning_rate": 2.9093645358927435e-05,
+      "loss": 0.7178,
+      "step": 10522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.392664718617925,
+      "learning_rate": 2.908953104273536e-05,
+      "loss": 0.7876,
+      "step": 10523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.821318893301318,
+      "learning_rate": 2.9085416612733312e-05,
+      "loss": 1.0811,
+      "step": 10524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.848176570144475,
+      "learning_rate": 2.908130206903579e-05,
+      "loss": 1.2734,
+      "step": 10525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.828669711603196,
+      "learning_rate": 2.907718741175728e-05,
+      "loss": 1.707,
+      "step": 10526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3159539487144207,
+      "learning_rate": 2.907307264101231e-05,
+      "loss": 1.4297,
+      "step": 10527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.455824029303165,
+      "learning_rate": 2.906895775691538e-05,
+      "loss": 1.0137,
+      "step": 10528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6232481044605613,
+      "learning_rate": 2.906484275958102e-05,
+      "loss": 1.5293,
+      "step": 10529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.32692635665068,
+      "learning_rate": 2.906072764912373e-05,
+      "loss": 1.0625,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.876425182539868,
+      "learning_rate": 2.9056612425658053e-05,
+      "loss": 0.665,
+      "step": 10531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.0103489804979855,
+      "learning_rate": 2.90524970892985e-05,
+      "loss": 1.2891,
+      "step": 10532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.382869232002661,
+      "learning_rate": 2.90483816401596e-05,
+      "loss": 0.8682,
+      "step": 10533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5086275956761384,
+      "learning_rate": 2.904426607835589e-05,
+      "loss": 0.7993,
+      "step": 10534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.398773865613349,
+      "learning_rate": 2.904015040400191e-05,
+      "loss": 1.3359,
+      "step": 10535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.027279518919048,
+      "learning_rate": 2.9036034617212178e-05,
+      "loss": 1.3398,
+      "step": 10536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.424362447517324,
+      "learning_rate": 2.9031918718101246e-05,
+      "loss": 0.8926,
+      "step": 10537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.716828820414596,
+      "learning_rate": 2.9027802706783657e-05,
+      "loss": 0.8281,
+      "step": 10538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9576192189593995,
+      "learning_rate": 2.9023686583373966e-05,
+      "loss": 0.8896,
+      "step": 10539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.398690578500008,
+      "learning_rate": 2.901957034798671e-05,
+      "loss": 1.0757,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3431922757637134,
+      "learning_rate": 2.901545400073645e-05,
+      "loss": 1.1943,
+      "step": 10541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6295683935235954,
+      "learning_rate": 2.901133754173774e-05,
+      "loss": 1.4766,
+      "step": 10542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.677401532411468,
+      "learning_rate": 2.900722097110513e-05,
+      "loss": 1.0664,
+      "step": 10543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9386571674620328,
+      "learning_rate": 2.9003104288953202e-05,
+      "loss": 0.75,
+      "step": 10544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.305757022261719,
+      "learning_rate": 2.8998987495396505e-05,
+      "loss": 0.9834,
+      "step": 10545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.079450138105206,
+      "learning_rate": 2.8994870590549623e-05,
+      "loss": 1.2607,
+      "step": 10546
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.877362648795439,
+      "learning_rate": 2.899075357452711e-05,
+      "loss": 0.6782,
+      "step": 10547
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7144836626079423,
+      "learning_rate": 2.8986636447443555e-05,
+      "loss": 1.3223,
+      "step": 10548
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6516561748354497,
+      "learning_rate": 2.898251920941353e-05,
+      "loss": 0.9365,
+      "step": 10549
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.005978826812333,
+      "learning_rate": 2.8978401860551613e-05,
+      "loss": 0.6946,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0822871696826897,
+      "learning_rate": 2.8974284400972395e-05,
+      "loss": 1.1621,
+      "step": 10551
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.250217891709669,
+      "learning_rate": 2.8970166830790456e-05,
+      "loss": 0.8408,
+      "step": 10552
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.893049595326569,
+      "learning_rate": 2.896604915012039e-05,
+      "loss": 1.0723,
+      "step": 10553
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.416770018588362,
+      "learning_rate": 2.89619313590768e-05,
+      "loss": 1.01,
+      "step": 10554
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.387240238937954,
+      "learning_rate": 2.8957813457774262e-05,
+      "loss": 0.866,
+      "step": 10555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8838626427181704,
+      "learning_rate": 2.8953695446327396e-05,
+      "loss": 1.4258,
+      "step": 10556
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2546186726111177,
+      "learning_rate": 2.8949577324850785e-05,
+      "loss": 1.2695,
+      "step": 10557
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8411538146991084,
+      "learning_rate": 2.894545909345906e-05,
+      "loss": 0.9854,
+      "step": 10558
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5812582685509025,
+      "learning_rate": 2.89413407522668e-05,
+      "loss": 1.1562,
+      "step": 10559
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.2380807996113266,
+      "learning_rate": 2.8937222301388646e-05,
+      "loss": 1.1777,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.045976397865767,
+      "learning_rate": 2.893310374093918e-05,
+      "loss": 1.1445,
+      "step": 10561
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3119591776467905,
+      "learning_rate": 2.892898507103305e-05,
+      "loss": 0.4067,
+      "step": 10562
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9771772793101257,
+      "learning_rate": 2.8924866291784865e-05,
+      "loss": 0.8391,
+      "step": 10563
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.3179521296176855,
+      "learning_rate": 2.8920747403309247e-05,
+      "loss": 0.8086,
+      "step": 10564
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.462616238829904,
+      "learning_rate": 2.8916628405720823e-05,
+      "loss": 1.6465,
+      "step": 10565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0434225537526043,
+      "learning_rate": 2.891250929913422e-05,
+      "loss": 1.7207,
+      "step": 10566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.1607893873738035,
+      "learning_rate": 2.890839008366409e-05,
+      "loss": 0.9312,
+      "step": 10567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3602796179362473,
+      "learning_rate": 2.8904270759425046e-05,
+      "loss": 1.1875,
+      "step": 10568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1299993579607905,
+      "learning_rate": 2.8900151326531737e-05,
+      "loss": 1.2803,
+      "step": 10569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8568204386747382,
+      "learning_rate": 2.8896031785098797e-05,
+      "loss": 0.6924,
+      "step": 10570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.3031384126430625,
+      "learning_rate": 2.889191213524089e-05,
+      "loss": 1.3633,
+      "step": 10571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.95988973231741,
+      "learning_rate": 2.8887792377072647e-05,
+      "loss": 1.3701,
+      "step": 10572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.410326352526687,
+      "learning_rate": 2.888367251070873e-05,
+      "loss": 0.8271,
+      "step": 10573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2274894214223333,
+      "learning_rate": 2.8879552536263788e-05,
+      "loss": 1.7148,
+      "step": 10574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.69369790050701,
+      "learning_rate": 2.887543245385247e-05,
+      "loss": 1.4395,
+      "step": 10575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5529470318664798,
+      "learning_rate": 2.8871312263589455e-05,
+      "loss": 1.1621,
+      "step": 10576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0833487573688476,
+      "learning_rate": 2.886719196558939e-05,
+      "loss": 0.7852,
+      "step": 10577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.223303823431321,
+      "learning_rate": 2.886307155996694e-05,
+      "loss": 0.6211,
+      "step": 10578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.499144864685505,
+      "learning_rate": 2.885895104683679e-05,
+      "loss": 1.2061,
+      "step": 10579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6777513810945663,
+      "learning_rate": 2.8854830426313606e-05,
+      "loss": 0.9888,
+      "step": 10580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.372646570427433,
+      "learning_rate": 2.8850709698512052e-05,
+      "loss": 1.1045,
+      "step": 10581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.054121265799742,
+      "learning_rate": 2.8846588863546827e-05,
+      "loss": 1.9043,
+      "step": 10582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3593538801245852,
+      "learning_rate": 2.884246792153259e-05,
+      "loss": 0.7542,
+      "step": 10583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.315098013601214,
+      "learning_rate": 2.8838346872584032e-05,
+      "loss": 0.6904,
+      "step": 10584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.706316260392238,
+      "learning_rate": 2.883422571681585e-05,
+      "loss": 0.9375,
+      "step": 10585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.66682390674717,
+      "learning_rate": 2.8830104454342728e-05,
+      "loss": 1.251,
+      "step": 10586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7152324151306586,
+      "learning_rate": 2.882598308527935e-05,
+      "loss": 1.1155,
+      "step": 10587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.033789822142252,
+      "learning_rate": 2.882186160974043e-05,
+      "loss": 1.0488,
+      "step": 10588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5277992875318676,
+      "learning_rate": 2.8817740027840644e-05,
+      "loss": 1.2803,
+      "step": 10589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.497756046309611,
+      "learning_rate": 2.8813618339694724e-05,
+      "loss": 0.8799,
+      "step": 10590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.042538891238769,
+      "learning_rate": 2.8809496545417347e-05,
+      "loss": 0.5557,
+      "step": 10591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1159302754659817,
+      "learning_rate": 2.880537464512324e-05,
+      "loss": 0.771,
+      "step": 10592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.000381671726815,
+      "learning_rate": 2.88012526389271e-05,
+      "loss": 1.3633,
+      "step": 10593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.81393895955636,
+      "learning_rate": 2.8797130526943643e-05,
+      "loss": 0.9092,
+      "step": 10594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.692741210461043,
+      "learning_rate": 2.87930083092876e-05,
+      "loss": 0.9326,
+      "step": 10595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.618737594732368,
+      "learning_rate": 2.878888598607367e-05,
+      "loss": 0.8223,
+      "step": 10596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.999851936581684,
+      "learning_rate": 2.878476355741659e-05,
+      "loss": 0.5713,
+      "step": 10597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.139485673883545,
+      "learning_rate": 2.8780641023431082e-05,
+      "loss": 1.0908,
+      "step": 10598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4222355727995364,
+      "learning_rate": 2.8776518384231872e-05,
+      "loss": 1.4561,
+      "step": 10599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.594236029622705,
+      "learning_rate": 2.8772395639933692e-05,
+      "loss": 0.8955,
+      "step": 10600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.718781511099798,
+      "learning_rate": 2.876827279065128e-05,
+      "loss": 1.292,
+      "step": 10601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.522523250657135,
+      "learning_rate": 2.8764149836499366e-05,
+      "loss": 1.0869,
+      "step": 10602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.40257921917686,
+      "learning_rate": 2.8760026777592695e-05,
+      "loss": 1.0879,
+      "step": 10603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.29734169788999,
+      "learning_rate": 2.875590361404601e-05,
+      "loss": 1.0938,
+      "step": 10604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.096212049765139,
+      "learning_rate": 2.875178034597406e-05,
+      "loss": 0.8408,
+      "step": 10605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8162845156705254,
+      "learning_rate": 2.8747656973491587e-05,
+      "loss": 1.0986,
+      "step": 10606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.165016102529828,
+      "learning_rate": 2.8743533496713345e-05,
+      "loss": 0.6997,
+      "step": 10607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.83096238584873,
+      "learning_rate": 2.8739409915754088e-05,
+      "loss": 1.2402,
+      "step": 10608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3915482277957105,
+      "learning_rate": 2.8735286230728587e-05,
+      "loss": 1.0479,
+      "step": 10609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.882629758652431,
+      "learning_rate": 2.873116244175158e-05,
+      "loss": 0.6807,
+      "step": 10610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.392362187534128,
+      "learning_rate": 2.8727038548937847e-05,
+      "loss": 1.2456,
+      "step": 10611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.222859666125543,
+      "learning_rate": 2.8722914552402143e-05,
+      "loss": 1.1465,
+      "step": 10612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7249649745545037,
+      "learning_rate": 2.8718790452259242e-05,
+      "loss": 1.623,
+      "step": 10613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.517531085930972,
+      "learning_rate": 2.871466624862392e-05,
+      "loss": 1.166,
+      "step": 10614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9839370724904377,
+      "learning_rate": 2.8710541941610953e-05,
+      "loss": 1.5703,
+      "step": 10615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3428393398411695,
+      "learning_rate": 2.8706417531335104e-05,
+      "loss": 1.2402,
+      "step": 10616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.210631656048919,
+      "learning_rate": 2.8702293017911168e-05,
+      "loss": 1.3574,
+      "step": 10617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.386577654885283,
+      "learning_rate": 2.869816840145393e-05,
+      "loss": 1.0869,
+      "step": 10618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.735746178810388,
+      "learning_rate": 2.8694043682078164e-05,
+      "loss": 1.1133,
+      "step": 10619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.009746360082191,
+      "learning_rate": 2.8689918859898673e-05,
+      "loss": 0.8809,
+      "step": 10620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.027380134609982,
+      "learning_rate": 2.8685793935030238e-05,
+      "loss": 0.9377,
+      "step": 10621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3870781410956736,
+      "learning_rate": 2.8681668907587654e-05,
+      "loss": 1.1406,
+      "step": 10622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7521781964717107,
+      "learning_rate": 2.867754377768573e-05,
+      "loss": 1.1438,
+      "step": 10623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4068626762731293,
+      "learning_rate": 2.867341854543926e-05,
+      "loss": 1.0381,
+      "step": 10624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6317322140001655,
+      "learning_rate": 2.8669293210963043e-05,
+      "loss": 0.6904,
+      "step": 10625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2526159762230984,
+      "learning_rate": 2.8665167774371894e-05,
+      "loss": 1.0405,
+      "step": 10626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.642965419495435,
+      "learning_rate": 2.866104223578061e-05,
+      "loss": 0.7837,
+      "step": 10627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.966888712163161,
+      "learning_rate": 2.8656916595304022e-05,
+      "loss": 1.6973,
+      "step": 10628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2471927843444606,
+      "learning_rate": 2.8652790853056927e-05,
+      "loss": 1.2266,
+      "step": 10629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2209997457120254,
+      "learning_rate": 2.8648665009154146e-05,
+      "loss": 1.457,
+      "step": 10630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.226243121063336,
+      "learning_rate": 2.864453906371051e-05,
+      "loss": 0.5674,
+      "step": 10631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7981077526592615,
+      "learning_rate": 2.8640413016840834e-05,
+      "loss": 0.8579,
+      "step": 10632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2560086023423698,
+      "learning_rate": 2.8636286868659952e-05,
+      "loss": 1.1143,
+      "step": 10633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.313029872914132,
+      "learning_rate": 2.8632160619282678e-05,
+      "loss": 1.001,
+      "step": 10634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.643519062750585,
+      "learning_rate": 2.8628034268823857e-05,
+      "loss": 1.1289,
+      "step": 10635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.9227985787108555,
+      "learning_rate": 2.8623907817398308e-05,
+      "loss": 0.8711,
+      "step": 10636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.289143098248063,
+      "learning_rate": 2.8619781265120894e-05,
+      "loss": 1.168,
+      "step": 10637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3743932852526726,
+      "learning_rate": 2.8615654612106433e-05,
+      "loss": 1.2246,
+      "step": 10638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1673681028509244,
+      "learning_rate": 2.861152785846979e-05,
+      "loss": 0.5571,
+      "step": 10639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.5621379551881125,
+      "learning_rate": 2.8607401004325773e-05,
+      "loss": 1.1631,
+      "step": 10640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.2165238552044135,
+      "learning_rate": 2.8603274049789274e-05,
+      "loss": 2.1973,
+      "step": 10641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.200531593513662,
+      "learning_rate": 2.8599146994975117e-05,
+      "loss": 1.0537,
+      "step": 10642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9984330356956648,
+      "learning_rate": 2.8595019839998168e-05,
+      "loss": 1.0862,
+      "step": 10643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6831794430060696,
+      "learning_rate": 2.859089258497328e-05,
+      "loss": 1.2285,
+      "step": 10644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4312655607769926,
+      "learning_rate": 2.8586765230015307e-05,
+      "loss": 1.3027,
+      "step": 10645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.845029103108469,
+      "learning_rate": 2.858263777523912e-05,
+      "loss": 1.0044,
+      "step": 10646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.172581290248546,
+      "learning_rate": 2.857851022075959e-05,
+      "loss": 1.1328,
+      "step": 10647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6327882307162214,
+      "learning_rate": 2.857438256669157e-05,
+      "loss": 1.1348,
+      "step": 10648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4368956379042213,
+      "learning_rate": 2.857025481314994e-05,
+      "loss": 1.0166,
+      "step": 10649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8566126960249572e-05,
+      "loss": 1.2461,
+      "step": 10650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.094316516144667,
+      "learning_rate": 2.8561999008105346e-05,
+      "loss": 1.0908,
+      "step": 10651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7416168606055034,
+      "learning_rate": 2.8557870956832132e-05,
+      "loss": 1.665,
+      "step": 10652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.447736032733841,
+      "learning_rate": 2.855374280654482e-05,
+      "loss": 0.6519,
+      "step": 10653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.120857933493004,
+      "learning_rate": 2.854961455735829e-05,
+      "loss": 1.1367,
+      "step": 10654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.036251543090981,
+      "learning_rate": 2.8545486209387434e-05,
+      "loss": 1.3506,
+      "step": 10655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.873427085173061,
+      "learning_rate": 2.8541357762747145e-05,
+      "loss": 0.7119,
+      "step": 10656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0626952439955195,
+      "learning_rate": 2.8537229217552302e-05,
+      "loss": 1.1855,
+      "step": 10657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2571495693272543,
+      "learning_rate": 2.8533100573917815e-05,
+      "loss": 1.2227,
+      "step": 10658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.6107662738482835,
+      "learning_rate": 2.8528971831958578e-05,
+      "loss": 1.1133,
+      "step": 10659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.187272192874716,
+      "learning_rate": 2.8524842991789492e-05,
+      "loss": 1.4033,
+      "step": 10660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.021450463143199,
+      "learning_rate": 2.852071405352546e-05,
+      "loss": 1.2617,
+      "step": 10661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6756603352869512,
+      "learning_rate": 2.851658501728139e-05,
+      "loss": 0.9375,
+      "step": 10662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.62447766627898,
+      "learning_rate": 2.8512455883172186e-05,
+      "loss": 1.8242,
+      "step": 10663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.932406615764714,
+      "learning_rate": 2.8508326651312768e-05,
+      "loss": 1.25,
+      "step": 10664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.169745083708397,
+      "learning_rate": 2.8504197321818045e-05,
+      "loss": 1.2432,
+      "step": 10665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.815089237674324,
+      "learning_rate": 2.850006789480294e-05,
+      "loss": 0.9541,
+      "step": 10666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.927810250890997,
+      "learning_rate": 2.8495938370382368e-05,
+      "loss": 1.1924,
+      "step": 10667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6645668424961464,
+      "learning_rate": 2.8491808748671255e-05,
+      "loss": 0.8457,
+      "step": 10668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1363443077507056,
+      "learning_rate": 2.8487679029784532e-05,
+      "loss": 1.1875,
+      "step": 10669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5036891420598444,
+      "learning_rate": 2.8483549213837113e-05,
+      "loss": 0.9443,
+      "step": 10670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.14602849810863,
+      "learning_rate": 2.8479419300943944e-05,
+      "loss": 0.874,
+      "step": 10671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6170631397732451,
+      "learning_rate": 2.847528929121994e-05,
+      "loss": 1.0098,
+      "step": 10672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6532432143403963,
+      "learning_rate": 2.8471159184780065e-05,
+      "loss": 0.6758,
+      "step": 10673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2464749219176383,
+      "learning_rate": 2.8467028981739236e-05,
+      "loss": 0.8926,
+      "step": 10674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.94227504184687,
+      "learning_rate": 2.8462898682212407e-05,
+      "loss": 1.041,
+      "step": 10675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.001570800001697,
+      "learning_rate": 2.8458768286314512e-05,
+      "loss": 1.082,
+      "step": 10676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.542407464638105,
+      "learning_rate": 2.8454637794160505e-05,
+      "loss": 0.7832,
+      "step": 10677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.304074377197216,
+      "learning_rate": 2.8450507205865328e-05,
+      "loss": 0.5181,
+      "step": 10678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.994043317533677,
+      "learning_rate": 2.8446376521543956e-05,
+      "loss": 0.8726,
+      "step": 10679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9493311593407987,
+      "learning_rate": 2.8442245741311313e-05,
+      "loss": 1.2393,
+      "step": 10680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6665417187527,
+      "learning_rate": 2.8438114865282377e-05,
+      "loss": 2.0586,
+      "step": 10681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.880807596201442,
+      "learning_rate": 2.8433983893572108e-05,
+      "loss": 0.8672,
+      "step": 10682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.500342098347387,
+      "learning_rate": 2.8429852826295462e-05,
+      "loss": 1.5234,
+      "step": 10683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.6656979963311755,
+      "learning_rate": 2.8425721663567412e-05,
+      "loss": 1.5034,
+      "step": 10684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0552127297366165,
+      "learning_rate": 2.8421590405502917e-05,
+      "loss": 0.5244,
+      "step": 10685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.4848385880257196,
+      "learning_rate": 2.8417459052216953e-05,
+      "loss": 1.3379,
+      "step": 10686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4680847044123704,
+      "learning_rate": 2.84133276038245e-05,
+      "loss": 0.9521,
+      "step": 10687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.753276096728802,
+      "learning_rate": 2.8409196060440535e-05,
+      "loss": 1.29,
+      "step": 10688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4360791130772417,
+      "learning_rate": 2.8405064422180017e-05,
+      "loss": 0.385,
+      "step": 10689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6790325821407763,
+      "learning_rate": 2.8400932689157954e-05,
+      "loss": 0.9951,
+      "step": 10690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.329758244034555,
+      "learning_rate": 2.8396800861489308e-05,
+      "loss": 1.1836,
+      "step": 10691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.292968083556022,
+      "learning_rate": 2.8392668939289085e-05,
+      "loss": 0.9668,
+      "step": 10692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.504674451464422,
+      "learning_rate": 2.8388536922672264e-05,
+      "loss": 1.0586,
+      "step": 10693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2309583671033155,
+      "learning_rate": 2.8384404811753845e-05,
+      "loss": 1.4316,
+      "step": 10694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8312085247439613,
+      "learning_rate": 2.8380272606648805e-05,
+      "loss": 1.3828,
+      "step": 10695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3640694949046113,
+      "learning_rate": 2.837614030747217e-05,
+      "loss": 0.3853,
+      "step": 10696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.119742076911301,
+      "learning_rate": 2.8372007914338915e-05,
+      "loss": 0.8838,
+      "step": 10697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4600052293861445,
+      "learning_rate": 2.836787542736406e-05,
+      "loss": 0.9551,
+      "step": 10698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4779890583956448,
+      "learning_rate": 2.8363742846662593e-05,
+      "loss": 0.8696,
+      "step": 10699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.116144915978091,
+      "learning_rate": 2.8359610172349537e-05,
+      "loss": 1.165,
+      "step": 10700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3484842931536094,
+      "learning_rate": 2.83554774045399e-05,
+      "loss": 0.9502,
+      "step": 10701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.294134880018454,
+      "learning_rate": 2.835134454334869e-05,
+      "loss": 1.2354,
+      "step": 10702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.866629314918549,
+      "learning_rate": 2.834721158889093e-05,
+      "loss": 1.0195,
+      "step": 10703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7853898911784896,
+      "learning_rate": 2.8343078541281632e-05,
+      "loss": 0.8906,
+      "step": 10704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.590152499588921,
+      "learning_rate": 2.8338945400635814e-05,
+      "loss": 0.5049,
+      "step": 10705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.049315199235998,
+      "learning_rate": 2.8334812167068502e-05,
+      "loss": 0.7375,
+      "step": 10706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.729622072463106,
+      "learning_rate": 2.8330678840694735e-05,
+      "loss": 1.1307,
+      "step": 10707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.766143922560318,
+      "learning_rate": 2.832654542162953e-05,
+      "loss": 0.7871,
+      "step": 10708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3008477018089017,
+      "learning_rate": 2.8322411909987917e-05,
+      "loss": 1.3633,
+      "step": 10709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.862800733303372,
+      "learning_rate": 2.8318278305884932e-05,
+      "loss": 1.6797,
+      "step": 10710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8055689256939262,
+      "learning_rate": 2.8314144609435616e-05,
+      "loss": 1.0332,
+      "step": 10711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.040615351430137,
+      "learning_rate": 2.8310010820755e-05,
+      "loss": 1.1191,
+      "step": 10712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.552395843993762,
+      "learning_rate": 2.8305876939958144e-05,
+      "loss": 1.4648,
+      "step": 10713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.101601703693905,
+      "learning_rate": 2.830174296716006e-05,
+      "loss": 1.1416,
+      "step": 10714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.403187871738058,
+      "learning_rate": 2.8297608902475824e-05,
+      "loss": 0.7268,
+      "step": 10715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5575462468385624,
+      "learning_rate": 2.8293474746020472e-05,
+      "loss": 1.0908,
+      "step": 10716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5065325672438297,
+      "learning_rate": 2.8289340497909063e-05,
+      "loss": 0.3812,
+      "step": 10717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2224677731695235,
+      "learning_rate": 2.828520615825664e-05,
+      "loss": 0.7588,
+      "step": 10718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.32489907031179,
+      "learning_rate": 2.8281071727178266e-05,
+      "loss": 0.9922,
+      "step": 10719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3081802023580433,
+      "learning_rate": 2.827693720478901e-05,
+      "loss": 0.8386,
+      "step": 10720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5558275951034597,
+      "learning_rate": 2.8272802591203922e-05,
+      "loss": 1.1055,
+      "step": 10721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.689865532677841,
+      "learning_rate": 2.8268667886538075e-05,
+      "loss": 1.2217,
+      "step": 10722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2159848974736818,
+      "learning_rate": 2.8264533090906514e-05,
+      "loss": 1.2715,
+      "step": 10723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7162073456969584,
+      "learning_rate": 2.8260398204424344e-05,
+      "loss": 1.0596,
+      "step": 10724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.744658745014927,
+      "learning_rate": 2.8256263227206615e-05,
+      "loss": 1.5605,
+      "step": 10725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6117775030332373,
+      "learning_rate": 2.825212815936841e-05,
+      "loss": 0.637,
+      "step": 10726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.752724919984225,
+      "learning_rate": 2.8247993001024793e-05,
+      "loss": 1.0078,
+      "step": 10727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8619663549307008,
+      "learning_rate": 2.8243857752290853e-05,
+      "loss": 1.0508,
+      "step": 10728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8269921077397033,
+      "learning_rate": 2.8239722413281673e-05,
+      "loss": 0.8901,
+      "step": 10729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.005035307259463,
+      "learning_rate": 2.8235586984112344e-05,
+      "loss": 0.8154,
+      "step": 10730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9602101105352094,
+      "learning_rate": 2.823145146489794e-05,
+      "loss": 1.25,
+      "step": 10731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5447308617614706,
+      "learning_rate": 2.8227315855753562e-05,
+      "loss": 1.1406,
+      "step": 10732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2564327683748524,
+      "learning_rate": 2.8223180156794292e-05,
+      "loss": 1.0322,
+      "step": 10733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.071721097311643,
+      "learning_rate": 2.821904436813524e-05,
+      "loss": 1.623,
+      "step": 10734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.651980210841757,
+      "learning_rate": 2.8214908489891485e-05,
+      "loss": 1.3125,
+      "step": 10735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.567199685622734,
+      "learning_rate": 2.821077252217814e-05,
+      "loss": 1.1484,
+      "step": 10736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.47398352894585,
+      "learning_rate": 2.8206636465110298e-05,
+      "loss": 1.1406,
+      "step": 10737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4932807730100985,
+      "learning_rate": 2.8202500318803072e-05,
+      "loss": 1.1958,
+      "step": 10738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8936829230889876,
+      "learning_rate": 2.8198364083371565e-05,
+      "loss": 0.876,
+      "step": 10739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5927388621753025,
+      "learning_rate": 2.819422775893089e-05,
+      "loss": 1.1494,
+      "step": 10740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.570101449958299,
+      "learning_rate": 2.819009134559616e-05,
+      "loss": 1.5288,
+      "step": 10741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.742988335531601,
+      "learning_rate": 2.818595484348247e-05,
+      "loss": 1.1143,
+      "step": 10742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.707740591671567,
+      "learning_rate": 2.8181818252704973e-05,
+      "loss": 1.2109,
+      "step": 10743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.313408210307684,
+      "learning_rate": 2.8177681573378755e-05,
+      "loss": 1.2832,
+      "step": 10744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4811383162185385,
+      "learning_rate": 2.8173544805618963e-05,
+      "loss": 1.0518,
+      "step": 10745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.636045107089427,
+      "learning_rate": 2.81694079495407e-05,
+      "loss": 0.7744,
+      "step": 10746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.978366415320024,
+      "learning_rate": 2.8165271005259113e-05,
+      "loss": 1.2383,
+      "step": 10747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.227311286167859,
+      "learning_rate": 2.8161133972889318e-05,
+      "loss": 0.7236,
+      "step": 10748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.373685934187078,
+      "learning_rate": 2.8156996852546458e-05,
+      "loss": 1.3159,
+      "step": 10749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2673880195935774,
+      "learning_rate": 2.8152859644345658e-05,
+      "loss": 1.2539,
+      "step": 10750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.046194764302326,
+      "learning_rate": 2.8148722348402057e-05,
+      "loss": 0.7212,
+      "step": 10751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0935250915889565,
+      "learning_rate": 2.8144584964830796e-05,
+      "loss": 0.9561,
+      "step": 10752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0049495249266656,
+      "learning_rate": 2.8140447493747012e-05,
+      "loss": 0.9453,
+      "step": 10753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.736501335635801,
+      "learning_rate": 2.8136309935265854e-05,
+      "loss": 1.2812,
+      "step": 10754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.499912832647851,
+      "learning_rate": 2.8132172289502463e-05,
+      "loss": 1.1289,
+      "step": 10755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.894075221330425,
+      "learning_rate": 2.8128034556572002e-05,
+      "loss": 1.0713,
+      "step": 10756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5942321696856534,
+      "learning_rate": 2.8123896736589605e-05,
+      "loss": 0.8164,
+      "step": 10757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.313127381046021,
+      "learning_rate": 2.811975882967044e-05,
+      "loss": 1.2715,
+      "step": 10758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.141361861549988,
+      "learning_rate": 2.811562083592965e-05,
+      "loss": 0.9893,
+      "step": 10759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.353443554178571,
+      "learning_rate": 2.8111482755482405e-05,
+      "loss": 0.6675,
+      "step": 10760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.633535602097314,
+      "learning_rate": 2.8107344588443856e-05,
+      "loss": 1.1309,
+      "step": 10761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0518494939133833,
+      "learning_rate": 2.8103206334929182e-05,
+      "loss": 0.7002,
+      "step": 10762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4838299423276875,
+      "learning_rate": 2.809906799505353e-05,
+      "loss": 1.0986,
+      "step": 10763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.956950253906342,
+      "learning_rate": 2.809492956893209e-05,
+      "loss": 1.1309,
+      "step": 10764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.840825348053289,
+      "learning_rate": 2.809079105668e-05,
+      "loss": 1.0715,
+      "step": 10765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.7006835683900094,
+      "learning_rate": 2.8086652458412464e-05,
+      "loss": 0.9932,
+      "step": 10766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.096727475210158,
+      "learning_rate": 2.8082513774244645e-05,
+      "loss": 1.6855,
+      "step": 10767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.290678012634617,
+      "learning_rate": 2.807837500429173e-05,
+      "loss": 1.1768,
+      "step": 10768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1499546834925485,
+      "learning_rate": 2.8074236148668887e-05,
+      "loss": 1.084,
+      "step": 10769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.838899805168825,
+      "learning_rate": 2.8070097207491297e-05,
+      "loss": 1.1123,
+      "step": 10770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.56289649657946,
+      "learning_rate": 2.8065958180874162e-05,
+      "loss": 1.5156,
+      "step": 10771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.729650853156391,
+      "learning_rate": 2.8061819068932654e-05,
+      "loss": 1.123,
+      "step": 10772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.127289570112163,
+      "learning_rate": 2.8057679871781966e-05,
+      "loss": 1.1543,
+      "step": 10773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8146533825600915,
+      "learning_rate": 2.805354058953729e-05,
+      "loss": 0.8982,
+      "step": 10774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.613682336551994,
+      "learning_rate": 2.804940122231383e-05,
+      "loss": 1.5928,
+      "step": 10775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.727038235409257,
+      "learning_rate": 2.8045261770226772e-05,
+      "loss": 1.375,
+      "step": 10776
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7569352178368502,
+      "learning_rate": 2.804112223339132e-05,
+      "loss": 1.0356,
+      "step": 10777
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5556044493250343,
+      "learning_rate": 2.803698261192267e-05,
+      "loss": 1.0854,
+      "step": 10778
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.84000141278084,
+      "learning_rate": 2.803284290593603e-05,
+      "loss": 1.4395,
+      "step": 10779
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.838278497019036,
+      "learning_rate": 2.8028703115546613e-05,
+      "loss": 1.0732,
+      "step": 10780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.341248904883437,
+      "learning_rate": 2.8024563240869615e-05,
+      "loss": 1.1865,
+      "step": 10781
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.501231653088068,
+      "learning_rate": 2.802042328202026e-05,
+      "loss": 1.3857,
+      "step": 10782
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3862401104617725,
+      "learning_rate": 2.8016283239113743e-05,
+      "loss": 1.2656,
+      "step": 10783
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.838528880985874,
+      "learning_rate": 2.80121431122653e-05,
+      "loss": 0.7271,
+      "step": 10784
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8313538050200755,
+      "learning_rate": 2.8008002901590142e-05,
+      "loss": 0.9111,
+      "step": 10785
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5392430637419245,
+      "learning_rate": 2.8003862607203484e-05,
+      "loss": 0.5078,
+      "step": 10786
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4318955944959915,
+      "learning_rate": 2.799972222922056e-05,
+      "loss": 0.6943,
+      "step": 10787
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.99503168267354,
+      "learning_rate": 2.799558176775658e-05,
+      "loss": 0.6785,
+      "step": 10788
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0841547173309145,
+      "learning_rate": 2.799144122292678e-05,
+      "loss": 1.0898,
+      "step": 10789
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.859847641826802,
+      "learning_rate": 2.798730059484639e-05,
+      "loss": 0.3501,
+      "step": 10790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.651887219479723,
+      "learning_rate": 2.798315988363064e-05,
+      "loss": 1.0137,
+      "step": 10791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.087731412176693,
+      "learning_rate": 2.7979019089394766e-05,
+      "loss": 0.8779,
+      "step": 10792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.130954366187297,
+      "learning_rate": 2.7974878212253997e-05,
+      "loss": 1.0088,
+      "step": 10793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.245207557004372,
+      "learning_rate": 2.797073725232358e-05,
+      "loss": 1.1484,
+      "step": 10794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.055097547556452,
+      "learning_rate": 2.796659620971876e-05,
+      "loss": 1.2793,
+      "step": 10795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.895343490877505,
+      "learning_rate": 2.7962455084554778e-05,
+      "loss": 1.2773,
+      "step": 10796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7058384677128258,
+      "learning_rate": 2.7958313876946863e-05,
+      "loss": 0.8164,
+      "step": 10797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.621994691734287,
+      "learning_rate": 2.795417258701029e-05,
+      "loss": 0.5996,
+      "step": 10798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0477279167339786,
+      "learning_rate": 2.7950031214860288e-05,
+      "loss": 0.7741,
+      "step": 10799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6328102360837353,
+      "learning_rate": 2.794588976061213e-05,
+      "loss": 1.3496,
+      "step": 10800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4518193943662188,
+      "learning_rate": 2.794174822438105e-05,
+      "loss": 0.7153,
+      "step": 10801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.387358718046273,
+      "learning_rate": 2.793760660628232e-05,
+      "loss": 0.8213,
+      "step": 10802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.6082784701984965,
+      "learning_rate": 2.7933464906431185e-05,
+      "loss": 1.5059,
+      "step": 10803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.851145262160755,
+      "learning_rate": 2.7929323124942923e-05,
+      "loss": 1.3193,
+      "step": 10804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.072160237633415,
+      "learning_rate": 2.7925181261932792e-05,
+      "loss": 1.6855,
+      "step": 10805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.143391068107616,
+      "learning_rate": 2.7921039317516052e-05,
+      "loss": 1.0371,
+      "step": 10806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.172291643132791,
+      "learning_rate": 2.7916897291807982e-05,
+      "loss": 0.8789,
+      "step": 10807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.847500518182535,
+      "learning_rate": 2.7912755184923845e-05,
+      "loss": 1.1494,
+      "step": 10808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.48181270540646,
+      "learning_rate": 2.7908612996978923e-05,
+      "loss": 1.0303,
+      "step": 10809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7466036891507204,
+      "learning_rate": 2.790447072808848e-05,
+      "loss": 1.4629,
+      "step": 10810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.025822827861018,
+      "learning_rate": 2.7900328378367796e-05,
+      "loss": 0.5469,
+      "step": 10811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.887208146047159,
+      "learning_rate": 2.789618594793215e-05,
+      "loss": 0.876,
+      "step": 10812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5647293610855852,
+      "learning_rate": 2.789204343689684e-05,
+      "loss": 1.6797,
+      "step": 10813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.600908351827906,
+      "learning_rate": 2.7887900845377134e-05,
+      "loss": 1.0586,
+      "step": 10814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0004319833637276,
+      "learning_rate": 2.7883758173488327e-05,
+      "loss": 0.8169,
+      "step": 10815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8113650524355767,
+      "learning_rate": 2.7879615421345695e-05,
+      "loss": 0.9873,
+      "step": 10816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.2281704959844175,
+      "learning_rate": 2.7875472589064545e-05,
+      "loss": 1.3496,
+      "step": 10817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9649891371632533,
+      "learning_rate": 2.787132967676016e-05,
+      "loss": 1.3467,
+      "step": 10818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4515357245204186,
+      "learning_rate": 2.786718668454784e-05,
+      "loss": 1.5352,
+      "step": 10819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3984662281018196,
+      "learning_rate": 2.7863043612542882e-05,
+      "loss": 0.8359,
+      "step": 10820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2272873083568014,
+      "learning_rate": 2.7858900460860586e-05,
+      "loss": 0.9192,
+      "step": 10821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.365221774810317,
+      "learning_rate": 2.785475722961625e-05,
+      "loss": 1.3691,
+      "step": 10822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.29111874733848,
+      "learning_rate": 2.7850613918925193e-05,
+      "loss": 1.1309,
+      "step": 10823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.432080249996296,
+      "learning_rate": 2.7846470528902703e-05,
+      "loss": 0.9541,
+      "step": 10824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.564817284294001,
+      "learning_rate": 2.7842327059664096e-05,
+      "loss": 1.4004,
+      "step": 10825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.002827296105977,
+      "learning_rate": 2.783818351132469e-05,
+      "loss": 0.835,
+      "step": 10826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.14778079763532,
+      "learning_rate": 2.783403988399979e-05,
+      "loss": 1.5488,
+      "step": 10827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.63405937208467,
+      "learning_rate": 2.7829896177804716e-05,
+      "loss": 0.7197,
+      "step": 10828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0725949930800702,
+      "learning_rate": 2.7825752392854775e-05,
+      "loss": 1.2368,
+      "step": 10829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.39548064137541,
+      "learning_rate": 2.78216085292653e-05,
+      "loss": 1.6543,
+      "step": 10830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.866174170601065,
+      "learning_rate": 2.781746458715161e-05,
+      "loss": 0.8506,
+      "step": 10831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7988864762159227,
+      "learning_rate": 2.781332056662903e-05,
+      "loss": 0.9663,
+      "step": 10832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.240454215245094,
+      "learning_rate": 2.7809176467812874e-05,
+      "loss": 1.1211,
+      "step": 10833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.974512839292241,
+      "learning_rate": 2.780503229081849e-05,
+      "loss": 0.4875,
+      "step": 10834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2763972466577673,
+      "learning_rate": 2.780088803576119e-05,
+      "loss": 1.0781,
+      "step": 10835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6542228592973136,
+      "learning_rate": 2.779674370275633e-05,
+      "loss": 1.2471,
+      "step": 10836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5730626972632624,
+      "learning_rate": 2.779259929191922e-05,
+      "loss": 1.0337,
+      "step": 10837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.9881887165947605,
+      "learning_rate": 2.7788454803365215e-05,
+      "loss": 0.6982,
+      "step": 10838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.524188172778125,
+      "learning_rate": 2.7784310237209644e-05,
+      "loss": 0.9551,
+      "step": 10839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.147003017382351,
+      "learning_rate": 2.778016559356785e-05,
+      "loss": 1.9191,
+      "step": 10840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7275617020128715,
+      "learning_rate": 2.7776020872555186e-05,
+      "loss": 1.7852,
+      "step": 10841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.297467149479964,
+      "learning_rate": 2.777187607428699e-05,
+      "loss": 0.7982,
+      "step": 10842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4342170517184183,
+      "learning_rate": 2.7767731198878606e-05,
+      "loss": 1.6211,
+      "step": 10843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.928812282164462,
+      "learning_rate": 2.7763586246445396e-05,
+      "loss": 1.2812,
+      "step": 10844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.006606279775972,
+      "learning_rate": 2.7759441217102706e-05,
+      "loss": 0.8867,
+      "step": 10845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1912078235437558,
+      "learning_rate": 2.7755296110965884e-05,
+      "loss": 1.0244,
+      "step": 10846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.053872204731536,
+      "learning_rate": 2.77511509281503e-05,
+      "loss": 1.0781,
+      "step": 10847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3876121249064917,
+      "learning_rate": 2.7747005668771293e-05,
+      "loss": 0.7539,
+      "step": 10848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0596815456029285,
+      "learning_rate": 2.774286033294425e-05,
+      "loss": 1.127,
+      "step": 10849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2467085970991514,
+      "learning_rate": 2.7738714920784513e-05,
+      "loss": 0.7998,
+      "step": 10850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.119612672617135,
+      "learning_rate": 2.773456943240746e-05,
+      "loss": 0.7588,
+      "step": 10851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1859013846892603,
+      "learning_rate": 2.773042386792845e-05,
+      "loss": 1.7305,
+      "step": 10852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.885201359322566,
+      "learning_rate": 2.7726278227462853e-05,
+      "loss": 0.7827,
+      "step": 10853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1523473803417756,
+      "learning_rate": 2.772213251112604e-05,
+      "loss": 1.293,
+      "step": 10854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.931321290314514,
+      "learning_rate": 2.77179867190334e-05,
+      "loss": 0.5995,
+      "step": 10855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.30916953611752,
+      "learning_rate": 2.7713840851300287e-05,
+      "loss": 0.9009,
+      "step": 10856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.426304660910528,
+      "learning_rate": 2.7709694908042088e-05,
+      "loss": 1.3906,
+      "step": 10857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2317262575250183,
+      "learning_rate": 2.770554888937419e-05,
+      "loss": 1.0605,
+      "step": 10858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1903785722489326,
+      "learning_rate": 2.7701402795411962e-05,
+      "loss": 1.252,
+      "step": 10859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.431495149661637,
+      "learning_rate": 2.7697256626270797e-05,
+      "loss": 1.0605,
+      "step": 10860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.762620748927508,
+      "learning_rate": 2.7693110382066074e-05,
+      "loss": 1.0088,
+      "step": 10861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3274820963399394,
+      "learning_rate": 2.7688964062913186e-05,
+      "loss": 0.6572,
+      "step": 10862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.561436036829783,
+      "learning_rate": 2.7684817668927522e-05,
+      "loss": 0.5776,
+      "step": 10863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.628487767641522,
+      "learning_rate": 2.7680671200224485e-05,
+      "loss": 0.7139,
+      "step": 10864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.413816039930164,
+      "learning_rate": 2.7676524656919445e-05,
+      "loss": 1.5156,
+      "step": 10865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.403074411623966,
+      "learning_rate": 2.767237803912783e-05,
+      "loss": 0.9844,
+      "step": 10866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.592534449969591,
+      "learning_rate": 2.7668231346965002e-05,
+      "loss": 0.75,
+      "step": 10867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.587453042603359,
+      "learning_rate": 2.7664084580546395e-05,
+      "loss": 1.2969,
+      "step": 10868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.082100185376727,
+      "learning_rate": 2.7659937739987397e-05,
+      "loss": 1.0605,
+      "step": 10869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8583607360000824,
+      "learning_rate": 2.7655790825403415e-05,
+      "loss": 1.0352,
+      "step": 10870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.289238516123563,
+      "learning_rate": 2.7651643836909845e-05,
+      "loss": 0.5889,
+      "step": 10871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.105571164535697,
+      "learning_rate": 2.764749677462212e-05,
+      "loss": 0.75,
+      "step": 10872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.376034962868895,
+      "learning_rate": 2.7643349638655636e-05,
+      "loss": 0.916,
+      "step": 10873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8338905773612821,
+      "learning_rate": 2.7639202429125804e-05,
+      "loss": 0.7423,
+      "step": 10874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.68963750956543,
+      "learning_rate": 2.7635055146148043e-05,
+      "loss": 0.9678,
+      "step": 10875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.279049269066226,
+      "learning_rate": 2.763090778983777e-05,
+      "loss": 0.7163,
+      "step": 10876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4283776386262557,
+      "learning_rate": 2.762676036031041e-05,
+      "loss": 0.4102,
+      "step": 10877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.542916152339831,
+      "learning_rate": 2.762261285768137e-05,
+      "loss": 0.9561,
+      "step": 10878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.896781418193122,
+      "learning_rate": 2.7618465282066092e-05,
+      "loss": 0.7334,
+      "step": 10879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1182274323090455,
+      "learning_rate": 2.7614317633579978e-05,
+      "loss": 0.8257,
+      "step": 10880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.1839990357202765,
+      "learning_rate": 2.7610169912338473e-05,
+      "loss": 1.0566,
+      "step": 10881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6846370085101183,
+      "learning_rate": 2.7606022118457002e-05,
+      "loss": 1.3086,
+      "step": 10882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.336587770720013,
+      "learning_rate": 2.7601874252051e-05,
+      "loss": 1.1367,
+      "step": 10883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.967527994999476,
+      "learning_rate": 2.75977263132359e-05,
+      "loss": 1.1484,
+      "step": 10884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7791817507889722,
+      "learning_rate": 2.7593578302127128e-05,
+      "loss": 0.6487,
+      "step": 10885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 28.34990049206798,
+      "learning_rate": 2.7589430218840123e-05,
+      "loss": 1.0908,
+      "step": 10886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.444719915330146,
+      "learning_rate": 2.7585282063490343e-05,
+      "loss": 1.123,
+      "step": 10887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.48990414094057,
+      "learning_rate": 2.7581133836193202e-05,
+      "loss": 1.0342,
+      "step": 10888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.247725820939775,
+      "learning_rate": 2.757698553706417e-05,
+      "loss": 0.6667,
+      "step": 10889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.612818007568381,
+      "learning_rate": 2.757283716621866e-05,
+      "loss": 1.0195,
+      "step": 10890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7794132070409745,
+      "learning_rate": 2.7568688723772157e-05,
+      "loss": 0.897,
+      "step": 10891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.950463575754109,
+      "learning_rate": 2.756454020984009e-05,
+      "loss": 1.1392,
+      "step": 10892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.979601492638175,
+      "learning_rate": 2.7560391624537914e-05,
+      "loss": 0.6143,
+      "step": 10893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4170282904500597,
+      "learning_rate": 2.755624296798107e-05,
+      "loss": 0.9893,
+      "step": 10894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.3733894241724744,
+      "learning_rate": 2.7552094240285036e-05,
+      "loss": 1.0723,
+      "step": 10895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.222165402511068,
+      "learning_rate": 2.754794544156526e-05,
+      "loss": 2.0605,
+      "step": 10896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.55282861116826,
+      "learning_rate": 2.7543796571937193e-05,
+      "loss": 0.8706,
+      "step": 10897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6432481185659604,
+      "learning_rate": 2.75396476315163e-05,
+      "loss": 0.5613,
+      "step": 10898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5680946456518763,
+      "learning_rate": 2.753549862041805e-05,
+      "loss": 0.8984,
+      "step": 10899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1573135548060183,
+      "learning_rate": 2.753134953875791e-05,
+      "loss": 1.1167,
+      "step": 10900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7889046411959355,
+      "learning_rate": 2.7527200386651336e-05,
+      "loss": 1.4043,
+      "step": 10901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1106302553065706,
+      "learning_rate": 2.7523051164213813e-05,
+      "loss": 1.0166,
+      "step": 10902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.277769440749893,
+      "learning_rate": 2.7518901871560797e-05,
+      "loss": 1.1855,
+      "step": 10903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.288525723480081,
+      "learning_rate": 2.7514752508807767e-05,
+      "loss": 0.6223,
+      "step": 10904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.793270384508254,
+      "learning_rate": 2.7510603076070195e-05,
+      "loss": 1.0527,
+      "step": 10905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.022682588550761,
+      "learning_rate": 2.7506453573463575e-05,
+      "loss": 0.6523,
+      "step": 10906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8719475321406915,
+      "learning_rate": 2.7502304001103362e-05,
+      "loss": 1.0967,
+      "step": 10907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.200629476385072,
+      "learning_rate": 2.749815435910505e-05,
+      "loss": 1.4902,
+      "step": 10908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.864876781876156,
+      "learning_rate": 2.749400464758412e-05,
+      "loss": 0.9399,
+      "step": 10909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.412234110830983,
+      "learning_rate": 2.748985486665605e-05,
+      "loss": 1.168,
+      "step": 10910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0771063121977247,
+      "learning_rate": 2.7485705016436336e-05,
+      "loss": 0.9136,
+      "step": 10911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9269182219396623,
+      "learning_rate": 2.7481555097040467e-05,
+      "loss": 0.8003,
+      "step": 10912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.163462862193689,
+      "learning_rate": 2.747740510858393e-05,
+      "loss": 0.6826,
+      "step": 10913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.7861283291972985,
+      "learning_rate": 2.7473255051182213e-05,
+      "loss": 1.1074,
+      "step": 10914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.652437896049913,
+      "learning_rate": 2.7469104924950815e-05,
+      "loss": 1.2656,
+      "step": 10915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9422900147516098,
+      "learning_rate": 2.746495473000524e-05,
+      "loss": 0.8474,
+      "step": 10916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2321204702072124,
+      "learning_rate": 2.746080446646097e-05,
+      "loss": 1.1484,
+      "step": 10917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2570214387199137,
+      "learning_rate": 2.7456654134433512e-05,
+      "loss": 1.2793,
+      "step": 10918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1540910167679796,
+      "learning_rate": 2.7452503734038376e-05,
+      "loss": 0.9941,
+      "step": 10919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.200563280647987,
+      "learning_rate": 2.7448353265391053e-05,
+      "loss": 0.6978,
+      "step": 10920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4638982491910952,
+      "learning_rate": 2.7444202728607065e-05,
+      "loss": 1.3281,
+      "step": 10921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2955362998968862,
+      "learning_rate": 2.7440052123801896e-05,
+      "loss": 1.2031,
+      "step": 10922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.228767088005052,
+      "learning_rate": 2.7435901451091078e-05,
+      "loss": 0.7349,
+      "step": 10923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0550242100981087,
+      "learning_rate": 2.7431750710590116e-05,
+      "loss": 0.6307,
+      "step": 10924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.202170774167292,
+      "learning_rate": 2.7427599902414525e-05,
+      "loss": 1.2656,
+      "step": 10925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6743915131962375,
+      "learning_rate": 2.742344902667981e-05,
+      "loss": 1.1309,
+      "step": 10926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.552533714998862,
+      "learning_rate": 2.7419298083501492e-05,
+      "loss": 0.8467,
+      "step": 10927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.77621685815935,
+      "learning_rate": 2.74151470729951e-05,
+      "loss": 1.2637,
+      "step": 10928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.775309549277095,
+      "learning_rate": 2.7410995995276146e-05,
+      "loss": 1.7363,
+      "step": 10929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6964712996692892,
+      "learning_rate": 2.7406844850460155e-05,
+      "loss": 1.3301,
+      "step": 10930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.090565848159626,
+      "learning_rate": 2.7402693638662658e-05,
+      "loss": 1.5078,
+      "step": 10931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.98079755740047,
+      "learning_rate": 2.7398542359999164e-05,
+      "loss": 1.3076,
+      "step": 10932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.741098998582024,
+      "learning_rate": 2.7394391014585218e-05,
+      "loss": 1.4941,
+      "step": 10933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.786321666372588,
+      "learning_rate": 2.739023960253635e-05,
+      "loss": 0.5618,
+      "step": 10934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.621305095344342,
+      "learning_rate": 2.738608812396808e-05,
+      "loss": 1.1719,
+      "step": 10935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7767693077096856,
+      "learning_rate": 2.7381936578995942e-05,
+      "loss": 1.6836,
+      "step": 10936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9317833712986783,
+      "learning_rate": 2.7377784967735488e-05,
+      "loss": 1.3105,
+      "step": 10937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.024555461027326,
+      "learning_rate": 2.737363329030225e-05,
+      "loss": 1.3799,
+      "step": 10938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2316670046809164,
+      "learning_rate": 2.736948154681176e-05,
+      "loss": 0.8652,
+      "step": 10939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8136252321772552,
+      "learning_rate": 2.7365329737379565e-05,
+      "loss": 0.6665,
+      "step": 10940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.573936442915297,
+      "learning_rate": 2.7361177862121195e-05,
+      "loss": 1.2524,
+      "step": 10941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.847066420729439,
+      "learning_rate": 2.7357025921152217e-05,
+      "loss": 1.2578,
+      "step": 10942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1972957509461892,
+      "learning_rate": 2.735287391458816e-05,
+      "loss": 1.3066,
+      "step": 10943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.389815816706191,
+      "learning_rate": 2.7348721842544596e-05,
+      "loss": 1.1152,
+      "step": 10944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.399874633602376,
+      "learning_rate": 2.734456970513704e-05,
+      "loss": 1.6069,
+      "step": 10945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6422815487736164,
+      "learning_rate": 2.7340417502481076e-05,
+      "loss": 1.2461,
+      "step": 10946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4008023907531,
+      "learning_rate": 2.7336265234692242e-05,
+      "loss": 1.0996,
+      "step": 10947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5203066083788372,
+      "learning_rate": 2.7332112901886098e-05,
+      "loss": 0.7051,
+      "step": 10948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8756011988735324,
+      "learning_rate": 2.7327960504178202e-05,
+      "loss": 0.9658,
+      "step": 10949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1104543679407417,
+      "learning_rate": 2.7323808041684107e-05,
+      "loss": 1.0078,
+      "step": 10950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.241347606291181,
+      "learning_rate": 2.731965551451939e-05,
+      "loss": 0.9604,
+      "step": 10951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7314437413407173,
+      "learning_rate": 2.7315502922799592e-05,
+      "loss": 1.1572,
+      "step": 10952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.244250728409606,
+      "learning_rate": 2.73113502666403e-05,
+      "loss": 0.7031,
+      "step": 10953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.367266882653156,
+      "learning_rate": 2.730719754615707e-05,
+      "loss": 1.7012,
+      "step": 10954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7665731238059177,
+      "learning_rate": 2.730304476146547e-05,
+      "loss": 1.3887,
+      "step": 10955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8506474947318883,
+      "learning_rate": 2.729889191268107e-05,
+      "loss": 0.6401,
+      "step": 10956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.428684825012922,
+      "learning_rate": 2.7294738999919445e-05,
+      "loss": 0.7544,
+      "step": 10957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5183482150183645,
+      "learning_rate": 2.7290586023296165e-05,
+      "loss": 1.0322,
+      "step": 10958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6671041984041914,
+      "learning_rate": 2.728643298292681e-05,
+      "loss": 1.3223,
+      "step": 10959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9241135126042432,
+      "learning_rate": 2.7282279878926957e-05,
+      "loss": 0.8428,
+      "step": 10960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.2449381813210065,
+      "learning_rate": 2.7278126711412188e-05,
+      "loss": 1.2656,
+      "step": 10961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.192990959117534,
+      "learning_rate": 2.7273973480498077e-05,
+      "loss": 1.3242,
+      "step": 10962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4365971189072155,
+      "learning_rate": 2.7269820186300217e-05,
+      "loss": 1.1592,
+      "step": 10963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.670110530176297,
+      "learning_rate": 2.7265666828934173e-05,
+      "loss": 0.9111,
+      "step": 10964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.744825607945361,
+      "learning_rate": 2.726151340851555e-05,
+      "loss": 0.8721,
+      "step": 10965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.274743879128819,
+      "learning_rate": 2.725735992515993e-05,
+      "loss": 2.2808,
+      "step": 10966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0712729157961887,
+      "learning_rate": 2.7253206378982898e-05,
+      "loss": 1.9766,
+      "step": 10967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1049172098365565,
+      "learning_rate": 2.7249052770100053e-05,
+      "loss": 0.5264,
+      "step": 10968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.94050844490555,
+      "learning_rate": 2.724489909862698e-05,
+      "loss": 1.5215,
+      "step": 10969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6721623734371014,
+      "learning_rate": 2.7240745364679286e-05,
+      "loss": 0.8906,
+      "step": 10970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0877202448898298,
+      "learning_rate": 2.7236591568372553e-05,
+      "loss": 0.6372,
+      "step": 10971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.428534471773249,
+      "learning_rate": 2.72324377098224e-05,
+      "loss": 0.7881,
+      "step": 10972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2732934305539008,
+      "learning_rate": 2.72282837891444e-05,
+      "loss": 1.0225,
+      "step": 10973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2663849306904864,
+      "learning_rate": 2.7224129806454176e-05,
+      "loss": 0.9746,
+      "step": 10974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.35067701935864,
+      "learning_rate": 2.7219975761867324e-05,
+      "loss": 0.5543,
+      "step": 10975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.93348416400232,
+      "learning_rate": 2.721582165549945e-05,
+      "loss": 1.6543,
+      "step": 10976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.031833022556942,
+      "learning_rate": 2.7211667487466162e-05,
+      "loss": 1.0908,
+      "step": 10977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4856503649888078,
+      "learning_rate": 2.7207513257883066e-05,
+      "loss": 0.519,
+      "step": 10978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7501046484014817,
+      "learning_rate": 2.7203358966865778e-05,
+      "loss": 1.0703,
+      "step": 10979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.055810434134093,
+      "learning_rate": 2.719920461452991e-05,
+      "loss": 1.1445,
+      "step": 10980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.383002455567723,
+      "learning_rate": 2.719505020099107e-05,
+      "loss": 1.1719,
+      "step": 10981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8802743712172996,
+      "learning_rate": 2.7190895726364875e-05,
+      "loss": 0.6709,
+      "step": 10982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0056800056349124,
+      "learning_rate": 2.718674119076694e-05,
+      "loss": 1.4102,
+      "step": 10983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3219286048561787,
+      "learning_rate": 2.7182586594312893e-05,
+      "loss": 1.3789,
+      "step": 10984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.400137916484213,
+      "learning_rate": 2.7178431937118352e-05,
+      "loss": 1.3555,
+      "step": 10985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.668666972737306,
+      "learning_rate": 2.7174277219298933e-05,
+      "loss": 0.9785,
+      "step": 10986
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.879273672844029,
+      "learning_rate": 2.7170122440970265e-05,
+      "loss": 0.7715,
+      "step": 10987
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.547750625413527,
+      "learning_rate": 2.7165967602247964e-05,
+      "loss": 1.1309,
+      "step": 10988
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.525096908992221,
+      "learning_rate": 2.7161812703247686e-05,
+      "loss": 0.6746,
+      "step": 10989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4372223916058666,
+      "learning_rate": 2.7157657744085025e-05,
+      "loss": 0.875,
+      "step": 10990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8633778929314317,
+      "learning_rate": 2.7153502724875633e-05,
+      "loss": 1.4961,
+      "step": 10991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0386035248993513,
+      "learning_rate": 2.7149347645735133e-05,
+      "loss": 0.6373,
+      "step": 10992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.127395513817469,
+      "learning_rate": 2.7145192506779172e-05,
+      "loss": 1.072,
+      "step": 10993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0226829040566865,
+      "learning_rate": 2.7141037308123367e-05,
+      "loss": 0.7983,
+      "step": 10994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.562061617464494,
+      "learning_rate": 2.713688204988337e-05,
+      "loss": 0.6108,
+      "step": 10995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.403700133950543,
+      "learning_rate": 2.7132726732174812e-05,
+      "loss": 1.1699,
+      "step": 10996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4633224226700494,
+      "learning_rate": 2.7128571355113342e-05,
+      "loss": 0.8208,
+      "step": 10997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7028193659347317,
+      "learning_rate": 2.7124415918814594e-05,
+      "loss": 0.7578,
+      "step": 10998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.843694136468318,
+      "learning_rate": 2.7120260423394222e-05,
+      "loss": 1.0171,
+      "step": 10999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0125881577884597,
+      "learning_rate": 2.7116104868967855e-05,
+      "loss": 0.7166,
+      "step": 11000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.155223415099846,
+      "learning_rate": 2.711194925565116e-05,
+      "loss": 0.6631,
+      "step": 11001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.401734000568489,
+      "learning_rate": 2.7107793583559777e-05,
+      "loss": 1.9219,
+      "step": 11002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3589667162776715,
+      "learning_rate": 2.710363785280935e-05,
+      "loss": 1.2383,
+      "step": 11003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1327082011113583,
+      "learning_rate": 2.7099482063515546e-05,
+      "loss": 1.3359,
+      "step": 11004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.293469664460616,
+      "learning_rate": 2.7095326215794002e-05,
+      "loss": 1.2949,
+      "step": 11005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8602901098158946,
+      "learning_rate": 2.709117030976039e-05,
+      "loss": 0.4706,
+      "step": 11006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.208139758952852,
+      "learning_rate": 2.7087014345530355e-05,
+      "loss": 0.6406,
+      "step": 11007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5260678008079995,
+      "learning_rate": 2.7082858323219563e-05,
+      "loss": 0.9031,
+      "step": 11008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.112304221137753,
+      "learning_rate": 2.707870224294367e-05,
+      "loss": 0.7029,
+      "step": 11009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.854463800114395,
+      "learning_rate": 2.7074546104818337e-05,
+      "loss": 0.7773,
+      "step": 11010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6951303614013757,
+      "learning_rate": 2.7070389908959232e-05,
+      "loss": 1.2617,
+      "step": 11011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.296597558514642,
+      "learning_rate": 2.7066233655482025e-05,
+      "loss": 1.541,
+      "step": 11012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0038519957933323,
+      "learning_rate": 2.7062077344502367e-05,
+      "loss": 1.0303,
+      "step": 11013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.24680836830892,
+      "learning_rate": 2.705792097613595e-05,
+      "loss": 1.2686,
+      "step": 11014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8159679648517035,
+      "learning_rate": 2.705376455049842e-05,
+      "loss": 1.2129,
+      "step": 11015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9023034563438348,
+      "learning_rate": 2.7049608067705456e-05,
+      "loss": 0.5337,
+      "step": 11016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.255526595740609,
+      "learning_rate": 2.704545152787274e-05,
+      "loss": 1.1875,
+      "step": 11017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5471404703717586,
+      "learning_rate": 2.7041294931115934e-05,
+      "loss": 1.2119,
+      "step": 11018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.337755909227414,
+      "learning_rate": 2.7037138277550728e-05,
+      "loss": 1.1318,
+      "step": 11019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.402011953872043,
+      "learning_rate": 2.7032981567292786e-05,
+      "loss": 1.0098,
+      "step": 11020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.208719711570988,
+      "learning_rate": 2.7028824800457804e-05,
+      "loss": 0.751,
+      "step": 11021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.8136052626201264,
+      "learning_rate": 2.702466797716145e-05,
+      "loss": 1.0596,
+      "step": 11022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5618725450424416,
+      "learning_rate": 2.7020511097519415e-05,
+      "loss": 1.2402,
+      "step": 11023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0602854292646193,
+      "learning_rate": 2.7016354161647363e-05,
+      "loss": 0.7832,
+      "step": 11024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.574991159516501,
+      "learning_rate": 2.7012197169661008e-05,
+      "loss": 1.1592,
+      "step": 11025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.353479224103549,
+      "learning_rate": 2.700804012167602e-05,
+      "loss": 1.0098,
+      "step": 11026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4670879618687174,
+      "learning_rate": 2.7003883017808103e-05,
+      "loss": 0.8208,
+      "step": 11027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.852204260404995,
+      "learning_rate": 2.6999725858172925e-05,
+      "loss": 1.3789,
+      "step": 11028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.97296798811173,
+      "learning_rate": 2.6995568642886195e-05,
+      "loss": 1.2129,
+      "step": 11029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.931462802425933,
+      "learning_rate": 2.6991411372063597e-05,
+      "loss": 1.1699,
+      "step": 11030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1022805047418567,
+      "learning_rate": 2.6987254045820843e-05,
+      "loss": 0.9651,
+      "step": 11031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9163714304862887,
+      "learning_rate": 2.698309666427361e-05,
+      "loss": 0.9941,
+      "step": 11032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7686565615901833,
+      "learning_rate": 2.6978939227537607e-05,
+      "loss": 0.8555,
+      "step": 11033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8205320455113068,
+      "learning_rate": 2.6974781735728528e-05,
+      "loss": 1.875,
+      "step": 11034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.437819017252197,
+      "learning_rate": 2.6970624188962078e-05,
+      "loss": 1.0752,
+      "step": 11035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7315055395911947,
+      "learning_rate": 2.696646658735396e-05,
+      "loss": 0.4463,
+      "step": 11036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.028585052733113,
+      "learning_rate": 2.6962308931019882e-05,
+      "loss": 0.8252,
+      "step": 11037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.105876140071343,
+      "learning_rate": 2.6958151220075533e-05,
+      "loss": 0.769,
+      "step": 11038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0868476394704354,
+      "learning_rate": 2.6953993454636638e-05,
+      "loss": 0.9634,
+      "step": 11039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.20171786603008,
+      "learning_rate": 2.6949835634818908e-05,
+      "loss": 1.1191,
+      "step": 11040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.050386285441147,
+      "learning_rate": 2.6945677760738043e-05,
+      "loss": 0.9116,
+      "step": 11041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.38665439901598,
+      "learning_rate": 2.6941519832509752e-05,
+      "loss": 0.6865,
+      "step": 11042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.563208784377182,
+      "learning_rate": 2.6937361850249755e-05,
+      "loss": 0.9639,
+      "step": 11043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.361366694461165,
+      "learning_rate": 2.6933203814073777e-05,
+      "loss": 0.8086,
+      "step": 11044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.840826761956187,
+      "learning_rate": 2.692904572409751e-05,
+      "loss": 1.418,
+      "step": 11045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.13208509075813,
+      "learning_rate": 2.6924887580436698e-05,
+      "loss": 1.0498,
+      "step": 11046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.128783322639913,
+      "learning_rate": 2.6920729383207033e-05,
+      "loss": 1.5186,
+      "step": 11047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.912359967635573,
+      "learning_rate": 2.6916571132524264e-05,
+      "loss": 1.0576,
+      "step": 11048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3724216720787488,
+      "learning_rate": 2.6912412828504098e-05,
+      "loss": 0.5042,
+      "step": 11049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4359682311608872,
+      "learning_rate": 2.6908254471262266e-05,
+      "loss": 1.0566,
+      "step": 11050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7097606907997998,
+      "learning_rate": 2.690409606091448e-05,
+      "loss": 1.0439,
+      "step": 11051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4790494447665155,
+      "learning_rate": 2.689993759757648e-05,
+      "loss": 1.2559,
+      "step": 11052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.243276377807321,
+      "learning_rate": 2.6895779081363993e-05,
+      "loss": 1.0332,
+      "step": 11053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3467779309634476,
+      "learning_rate": 2.689162051239274e-05,
+      "loss": 1.3047,
+      "step": 11054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.083104632032151,
+      "learning_rate": 2.6887461890778465e-05,
+      "loss": 1.0679,
+      "step": 11055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.830829739273434,
+      "learning_rate": 2.688330321663689e-05,
+      "loss": 1.1758,
+      "step": 11056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4788546089127688,
+      "learning_rate": 2.687914449008375e-05,
+      "loss": 0.7715,
+      "step": 11057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.86161159695091,
+      "learning_rate": 2.6874985711234784e-05,
+      "loss": 1.5146,
+      "step": 11058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.181366201093962,
+      "learning_rate": 2.6870826880205734e-05,
+      "loss": 1.3262,
+      "step": 11059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.708411705797056,
+      "learning_rate": 2.6866667997112333e-05,
+      "loss": 1.332,
+      "step": 11060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1929842574020904,
+      "learning_rate": 2.6862509062070317e-05,
+      "loss": 0.792,
+      "step": 11061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4876882663964555,
+      "learning_rate": 2.6858350075195438e-05,
+      "loss": 1.1333,
+      "step": 11062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1703179206632597,
+      "learning_rate": 2.685419103660343e-05,
+      "loss": 0.9854,
+      "step": 11063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1206931057220952,
+      "learning_rate": 2.6850031946410038e-05,
+      "loss": 0.9314,
+      "step": 11064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.449845959238871,
+      "learning_rate": 2.6845872804731014e-05,
+      "loss": 1.1768,
+      "step": 11065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.077953583487329,
+      "learning_rate": 2.68417136116821e-05,
+      "loss": 1.0669,
+      "step": 11066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.0016066206104,
+      "learning_rate": 2.6837554367379046e-05,
+      "loss": 1.7793,
+      "step": 11067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7942245114561506,
+      "learning_rate": 2.68333950719376e-05,
+      "loss": 0.6982,
+      "step": 11068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.727215783236397,
+      "learning_rate": 2.6829235725473524e-05,
+      "loss": 0.6699,
+      "step": 11069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.899625423631584,
+      "learning_rate": 2.6825076328102548e-05,
+      "loss": 0.7944,
+      "step": 11070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.009379756282042,
+      "learning_rate": 2.682091687994045e-05,
+      "loss": 2.709,
+      "step": 11071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.635600258283882,
+      "learning_rate": 2.6816757381102975e-05,
+      "loss": 0.8989,
+      "step": 11072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0440793346603496,
+      "learning_rate": 2.6812597831705883e-05,
+      "loss": 1.0312,
+      "step": 11073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.0808038306864365,
+      "learning_rate": 2.6808438231864935e-05,
+      "loss": 1.4463,
+      "step": 11074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4899838042224998,
+      "learning_rate": 2.6804278581695875e-05,
+      "loss": 1.5332,
+      "step": 11075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.483141610939507,
+      "learning_rate": 2.6800118881314485e-05,
+      "loss": 0.5612,
+      "step": 11076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2072728927795637,
+      "learning_rate": 2.6795959130836517e-05,
+      "loss": 1.2031,
+      "step": 11077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7047806913511496,
+      "learning_rate": 2.6791799330377744e-05,
+      "loss": 1.0605,
+      "step": 11078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.745325925394805,
+      "learning_rate": 2.6787639480053918e-05,
+      "loss": 1.2002,
+      "step": 11079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.571371576671863,
+      "learning_rate": 2.6783479579980807e-05,
+      "loss": 1.8242,
+      "step": 11080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3364184566659136,
+      "learning_rate": 2.6779319630274195e-05,
+      "loss": 0.7871,
+      "step": 11081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6162200358969394,
+      "learning_rate": 2.6775159631049844e-05,
+      "loss": 0.5791,
+      "step": 11082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0792858352095736,
+      "learning_rate": 2.6770999582423516e-05,
+      "loss": 1.5762,
+      "step": 11083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.548967681279182,
+      "learning_rate": 2.6766839484510997e-05,
+      "loss": 1.1816,
+      "step": 11084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.714276997587435,
+      "learning_rate": 2.676267933742805e-05,
+      "loss": 0.9241,
+      "step": 11085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3943033964991036,
+      "learning_rate": 2.6758519141290456e-05,
+      "loss": 1.1055,
+      "step": 11086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.649438017582526,
+      "learning_rate": 2.675435889621399e-05,
+      "loss": 1.2861,
+      "step": 11087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.69325249241073,
+      "learning_rate": 2.675019860231443e-05,
+      "loss": 1.1084,
+      "step": 11088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5040805892833666,
+      "learning_rate": 2.6746038259707557e-05,
+      "loss": 0.7563,
+      "step": 11089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.154005473266248,
+      "learning_rate": 2.6741877868509145e-05,
+      "loss": 1.2539,
+      "step": 11090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.43697293237316,
+      "learning_rate": 2.6737717428834986e-05,
+      "loss": 0.7424,
+      "step": 11091
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.177984194261212,
+      "learning_rate": 2.6733556940800857e-05,
+      "loss": 1.3848,
+      "step": 11092
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.166040475737218,
+      "learning_rate": 2.672939640452255e-05,
+      "loss": 0.7031,
+      "step": 11093
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4018991942903165,
+      "learning_rate": 2.6725235820115828e-05,
+      "loss": 0.9971,
+      "step": 11094
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.468621799533878,
+      "learning_rate": 2.6721075187696513e-05,
+      "loss": 2.1836,
+      "step": 11095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9713608354142043,
+      "learning_rate": 2.671691450738037e-05,
+      "loss": 0.8433,
+      "step": 11096
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.078224466567835,
+      "learning_rate": 2.6712753779283194e-05,
+      "loss": 1.1182,
+      "step": 11097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8242994227876923,
+      "learning_rate": 2.6708593003520775e-05,
+      "loss": 1.3691,
+      "step": 11098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.310627669068978,
+      "learning_rate": 2.6704432180208915e-05,
+      "loss": 1.0791,
+      "step": 11099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.020434866878648,
+      "learning_rate": 2.67002713094634e-05,
+      "loss": 1.3301,
+      "step": 11100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6872522994044115,
+      "learning_rate": 2.6696110391400024e-05,
+      "loss": 1.2222,
+      "step": 11101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.982380322551744,
+      "learning_rate": 2.6691949426134586e-05,
+      "loss": 0.5857,
+      "step": 11102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.800195693943093,
+      "learning_rate": 2.6687788413782882e-05,
+      "loss": 0.7979,
+      "step": 11103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.267437113302457,
+      "learning_rate": 2.668362735446072e-05,
+      "loss": 1.0977,
+      "step": 11104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.459203296760429,
+      "learning_rate": 2.667946624828389e-05,
+      "loss": 0.3987,
+      "step": 11105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4911832112098327,
+      "learning_rate": 2.6675305095368204e-05,
+      "loss": 1.332,
+      "step": 11106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9852514128503156,
+      "learning_rate": 2.6671143895829452e-05,
+      "loss": 1.1562,
+      "step": 11107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2616863874317312,
+      "learning_rate": 2.6666982649783444e-05,
+      "loss": 1.02,
+      "step": 11108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8211093104569045,
+      "learning_rate": 2.666282135734599e-05,
+      "loss": 1.4043,
+      "step": 11109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.19369289398813,
+      "learning_rate": 2.6658660018632902e-05,
+      "loss": 1.3223,
+      "step": 11110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.541921913074577,
+      "learning_rate": 2.6654498633759973e-05,
+      "loss": 0.8994,
+      "step": 11111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2850515959010345,
+      "learning_rate": 2.6650337202843024e-05,
+      "loss": 1.334,
+      "step": 11112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.702366540597106,
+      "learning_rate": 2.664617572599786e-05,
+      "loss": 1.1738,
+      "step": 11113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.102636810569329,
+      "learning_rate": 2.66420142033403e-05,
+      "loss": 0.854,
+      "step": 11114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.767601935482514,
+      "learning_rate": 2.6637852634986143e-05,
+      "loss": 0.7788,
+      "step": 11115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0779518686324936,
+      "learning_rate": 2.6633691021051226e-05,
+      "loss": 0.9814,
+      "step": 11116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.903371554479691,
+      "learning_rate": 2.662952936165134e-05,
+      "loss": 1.623,
+      "step": 11117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.804993934148802,
+      "learning_rate": 2.662536765690233e-05,
+      "loss": 0.4956,
+      "step": 11118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1357570345920265,
+      "learning_rate": 2.6621205906919988e-05,
+      "loss": 0.3457,
+      "step": 11119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9431109942980926,
+      "learning_rate": 2.6617044111820156e-05,
+      "loss": 0.4696,
+      "step": 11120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6865491772694123,
+      "learning_rate": 2.6612882271718637e-05,
+      "loss": 0.9912,
+      "step": 11121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.818179918057518,
+      "learning_rate": 2.6608720386731262e-05,
+      "loss": 1.4492,
+      "step": 11122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.726664893733461,
+      "learning_rate": 2.660455845697386e-05,
+      "loss": 0.7065,
+      "step": 11123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3643872151727963,
+      "learning_rate": 2.660039648256224e-05,
+      "loss": 1.2964,
+      "step": 11124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3555267293448368,
+      "learning_rate": 2.6596234463612246e-05,
+      "loss": 0.9072,
+      "step": 11125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.735871976615256,
+      "learning_rate": 2.6592072400239685e-05,
+      "loss": 1.1895,
+      "step": 11126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5004030856379384,
+      "learning_rate": 2.6587910292560407e-05,
+      "loss": 0.7317,
+      "step": 11127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0108316700840407,
+      "learning_rate": 2.658374814069023e-05,
+      "loss": 0.5679,
+      "step": 11128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6918215050661733,
+      "learning_rate": 2.657958594474499e-05,
+      "loss": 1.1943,
+      "step": 11129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9282379226324777,
+      "learning_rate": 2.6575423704840506e-05,
+      "loss": 1.2148,
+      "step": 11130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.633363479329034,
+      "learning_rate": 2.657126142109263e-05,
+      "loss": 1.1973,
+      "step": 11131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.162648433001557,
+      "learning_rate": 2.6567099093617187e-05,
+      "loss": 1.0908,
+      "step": 11132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3901789629342645,
+      "learning_rate": 2.656293672253001e-05,
+      "loss": 0.9824,
+      "step": 11133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.027426975587918,
+      "learning_rate": 2.655877430794695e-05,
+      "loss": 2.4434,
+      "step": 11134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0265711964322852,
+      "learning_rate": 2.6554611849983825e-05,
+      "loss": 0.4751,
+      "step": 11135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.388876839279765,
+      "learning_rate": 2.655044934875649e-05,
+      "loss": 1.4297,
+      "step": 11136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8907637279316165,
+      "learning_rate": 2.654628680438078e-05,
+      "loss": 0.9658,
+      "step": 11137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.204215270589947,
+      "learning_rate": 2.654212421697254e-05,
+      "loss": 0.918,
+      "step": 11138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.543074879577801,
+      "learning_rate": 2.6537961586647615e-05,
+      "loss": 1.1982,
+      "step": 11139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.029088715227533,
+      "learning_rate": 2.6533798913521833e-05,
+      "loss": 0.9697,
+      "step": 11140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4407426832597965,
+      "learning_rate": 2.6529636197711055e-05,
+      "loss": 1.5703,
+      "step": 11141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.2848820559495495,
+      "learning_rate": 2.6525473439331133e-05,
+      "loss": 0.6584,
+      "step": 11142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.585187132499509,
+      "learning_rate": 2.6521310638497898e-05,
+      "loss": 0.8663,
+      "step": 11143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4031703921426772,
+      "learning_rate": 2.6517147795327213e-05,
+      "loss": 0.9307,
+      "step": 11144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7224520328944517,
+      "learning_rate": 2.6512984909934913e-05,
+      "loss": 1.1465,
+      "step": 11145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.595726638813738,
+      "learning_rate": 2.650882198243687e-05,
+      "loss": 1.0088,
+      "step": 11146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.054828312374678,
+      "learning_rate": 2.650465901294892e-05,
+      "loss": 0.9863,
+      "step": 11147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2061502234182906,
+      "learning_rate": 2.6500496001586924e-05,
+      "loss": 1.0549,
+      "step": 11148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.365150910093191,
+      "learning_rate": 2.6496332948466728e-05,
+      "loss": 1.5264,
+      "step": 11149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6648740865980205,
+      "learning_rate": 2.64921698537042e-05,
+      "loss": 0.958,
+      "step": 11150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.556279424297483,
+      "learning_rate": 2.64880067174152e-05,
+      "loss": 0.835,
+      "step": 11151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.277646183719963,
+      "learning_rate": 2.648384353971557e-05,
+      "loss": 0.9473,
+      "step": 11152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2670290049618638,
+      "learning_rate": 2.6479680320721177e-05,
+      "loss": 1.0918,
+      "step": 11153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.399847695575688,
+      "learning_rate": 2.6475517060547888e-05,
+      "loss": 0.8208,
+      "step": 11154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8109748943867428,
+      "learning_rate": 2.6471353759311556e-05,
+      "loss": 1.084,
+      "step": 11155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1075702586252945,
+      "learning_rate": 2.646719041712806e-05,
+      "loss": 1.125,
+      "step": 11156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.638500068281875,
+      "learning_rate": 2.6463027034113237e-05,
+      "loss": 1.1426,
+      "step": 11157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.538186831836282,
+      "learning_rate": 2.6458863610382977e-05,
+      "loss": 1.1084,
+      "step": 11158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0536298945413987,
+      "learning_rate": 2.6454700146053128e-05,
+      "loss": 0.647,
+      "step": 11159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.699059639866623,
+      "learning_rate": 2.645053664123957e-05,
+      "loss": 1.1641,
+      "step": 11160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.596223051020816,
+      "learning_rate": 2.6446373096058175e-05,
+      "loss": 0.8418,
+      "step": 11161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3720240771018903,
+      "learning_rate": 2.6442209510624793e-05,
+      "loss": 1.1367,
+      "step": 11162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1060015143435606,
+      "learning_rate": 2.6438045885055314e-05,
+      "loss": 0.9023,
+      "step": 11163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8564155981360453,
+      "learning_rate": 2.64338822194656e-05,
+      "loss": 1.9355,
+      "step": 11164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.40331086985778,
+      "learning_rate": 2.6429718513971537e-05,
+      "loss": 0.9746,
+      "step": 11165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.627588539845804,
+      "learning_rate": 2.6425554768688977e-05,
+      "loss": 1.374,
+      "step": 11166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.50127310389554,
+      "learning_rate": 2.642139098373382e-05,
+      "loss": 0.9487,
+      "step": 11167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.547225912345779,
+      "learning_rate": 2.641722715922192e-05,
+      "loss": 0.7694,
+      "step": 11168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.026597196095205,
+      "learning_rate": 2.641306329526917e-05,
+      "loss": 0.8384,
+      "step": 11169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1539093803463265,
+      "learning_rate": 2.6408899391991443e-05,
+      "loss": 1.4688,
+      "step": 11170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.493803174626226,
+      "learning_rate": 2.6404735449504624e-05,
+      "loss": 0.9033,
+      "step": 11171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.434031635401691,
+      "learning_rate": 2.6400571467924585e-05,
+      "loss": 1.3301,
+      "step": 11172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7899255772432365,
+      "learning_rate": 2.639640744736721e-05,
+      "loss": 1.2441,
+      "step": 11173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5543152331255343,
+      "learning_rate": 2.639224338794839e-05,
+      "loss": 0.7632,
+      "step": 11174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9118948639345477,
+      "learning_rate": 2.6388079289783996e-05,
+      "loss": 1.1953,
+      "step": 11175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.049772166513289,
+      "learning_rate": 2.6383915152989923e-05,
+      "loss": 0.6187,
+      "step": 11176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.675028773603876,
+      "learning_rate": 2.6379750977682054e-05,
+      "loss": 0.9297,
+      "step": 11177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2125866592533954,
+      "learning_rate": 2.6375586763976285e-05,
+      "loss": 0.5723,
+      "step": 11178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1466314319222577,
+      "learning_rate": 2.6371422511988492e-05,
+      "loss": 0.8203,
+      "step": 11179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.651817752778538,
+      "learning_rate": 2.6367258221834567e-05,
+      "loss": 0.7764,
+      "step": 11180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.637023350159751,
+      "learning_rate": 2.6363093893630404e-05,
+      "loss": 1.2773,
+      "step": 11181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.73017550747231,
+      "learning_rate": 2.6358929527491888e-05,
+      "loss": 0.9766,
+      "step": 11182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.342505951567045,
+      "learning_rate": 2.635476512353492e-05,
+      "loss": 1.4961,
+      "step": 11183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.883724405644225,
+      "learning_rate": 2.635060068187539e-05,
+      "loss": 1.416,
+      "step": 11184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6436117762872002,
+      "learning_rate": 2.6346436202629194e-05,
+      "loss": 1.4102,
+      "step": 11185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.166363646029535,
+      "learning_rate": 2.6342271685912225e-05,
+      "loss": 1.1621,
+      "step": 11186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7136500087399043,
+      "learning_rate": 2.6338107131840377e-05,
+      "loss": 1.0977,
+      "step": 11187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.417870736793286,
+      "learning_rate": 2.6333942540529566e-05,
+      "loss": 0.6663,
+      "step": 11188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.268490367146707,
+      "learning_rate": 2.6329777912095666e-05,
+      "loss": 1.1621,
+      "step": 11189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5054117755936254,
+      "learning_rate": 2.632561324665459e-05,
+      "loss": 0.9043,
+      "step": 11190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.3236600452570055,
+      "learning_rate": 2.6321448544322237e-05,
+      "loss": 0.6509,
+      "step": 11191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.149729956607846,
+      "learning_rate": 2.631728380521451e-05,
+      "loss": 1.3535,
+      "step": 11192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.667461849765733,
+      "learning_rate": 2.6313119029447313e-05,
+      "loss": 0.4846,
+      "step": 11193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8035737271004237,
+      "learning_rate": 2.6308954217136543e-05,
+      "loss": 0.9404,
+      "step": 11194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.765045768235394,
+      "learning_rate": 2.6304789368398114e-05,
+      "loss": 1.7344,
+      "step": 11195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3781792292383295,
+      "learning_rate": 2.6300624483347926e-05,
+      "loss": 0.7129,
+      "step": 11196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2205359564594858,
+      "learning_rate": 2.6296459562101888e-05,
+      "loss": 1.459,
+      "step": 11197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.130551564567835,
+      "learning_rate": 2.629229460477591e-05,
+      "loss": 0.9668,
+      "step": 11198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8829317559280483,
+      "learning_rate": 2.62881296114859e-05,
+      "loss": 0.9712,
+      "step": 11199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8869637592840298,
+      "learning_rate": 2.6283964582347758e-05,
+      "loss": 1.0645,
+      "step": 11200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.829033805712605,
+      "learning_rate": 2.6279799517477414e-05,
+      "loss": 1.2139,
+      "step": 11201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.1723357596009025,
+      "learning_rate": 2.627563441699077e-05,
+      "loss": 1.2607,
+      "step": 11202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.846829668742965,
+      "learning_rate": 2.6271469281003742e-05,
+      "loss": 1.5859,
+      "step": 11203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.282743095998295,
+      "learning_rate": 2.6267304109632236e-05,
+      "loss": 0.8306,
+      "step": 11204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.280447557525336,
+      "learning_rate": 2.626313890299217e-05,
+      "loss": 1.3447,
+      "step": 11205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.453581258054931,
+      "learning_rate": 2.6258973661199465e-05,
+      "loss": 0.7517,
+      "step": 11206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.1736539333997795,
+      "learning_rate": 2.625480838437004e-05,
+      "loss": 1.3477,
+      "step": 11207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6471995833626285,
+      "learning_rate": 2.62506430726198e-05,
+      "loss": 1.5586,
+      "step": 11208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5277133211536786,
+      "learning_rate": 2.6246477726064683e-05,
+      "loss": 0.699,
+      "step": 11209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.9693649199389185,
+      "learning_rate": 2.624231234482059e-05,
+      "loss": 0.8721,
+      "step": 11210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.656172988561965,
+      "learning_rate": 2.6238146929003454e-05,
+      "loss": 0.9902,
+      "step": 11211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5157581897513532,
+      "learning_rate": 2.6233981478729196e-05,
+      "loss": 1.5215,
+      "step": 11212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.9188525410958315,
+      "learning_rate": 2.6229815994113732e-05,
+      "loss": 1.0137,
+      "step": 11213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1255256900172577,
+      "learning_rate": 2.622565047527299e-05,
+      "loss": 1.042,
+      "step": 11214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.112519019227496,
+      "learning_rate": 2.6221484922322886e-05,
+      "loss": 1.0879,
+      "step": 11215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.086592553542035,
+      "learning_rate": 2.621731933537937e-05,
+      "loss": 0.6992,
+      "step": 11216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.723020000503302,
+      "learning_rate": 2.6213153714558343e-05,
+      "loss": 0.7822,
+      "step": 11217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0528429320809387,
+      "learning_rate": 2.6208988059975746e-05,
+      "loss": 0.835,
+      "step": 11218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.876282572247227,
+      "learning_rate": 2.6204822371747494e-05,
+      "loss": 0.8828,
+      "step": 11219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7230933562325426,
+      "learning_rate": 2.6200656649989537e-05,
+      "loss": 1.0361,
+      "step": 11220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.426716352450163,
+      "learning_rate": 2.6196490894817787e-05,
+      "loss": 1.1777,
+      "step": 11221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.181433995507888,
+      "learning_rate": 2.6192325106348187e-05,
+      "loss": 0.7793,
+      "step": 11222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.109911782535288,
+      "learning_rate": 2.6188159284696663e-05,
+      "loss": 1.4531,
+      "step": 11223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3458242774187803,
+      "learning_rate": 2.6183993429979148e-05,
+      "loss": 1.1973,
+      "step": 11224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.505223871719058,
+      "learning_rate": 2.6179827542311574e-05,
+      "loss": 0.4736,
+      "step": 11225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.84791712357453,
+      "learning_rate": 2.6175661621809893e-05,
+      "loss": 0.79,
+      "step": 11226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1342461207786787,
+      "learning_rate": 2.617149566859002e-05,
+      "loss": 1.1953,
+      "step": 11227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8813500972414885,
+      "learning_rate": 2.6167329682767894e-05,
+      "loss": 1.0762,
+      "step": 11228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.569634994244795,
+      "learning_rate": 2.6163163664459467e-05,
+      "loss": 1.0146,
+      "step": 11229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.743351207005857,
+      "learning_rate": 2.6158997613780667e-05,
+      "loss": 1.4541,
+      "step": 11230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.665656875510785,
+      "learning_rate": 2.6154831530847434e-05,
+      "loss": 0.8154,
+      "step": 11231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.46023772447902,
+      "learning_rate": 2.6150665415775712e-05,
+      "loss": 1.001,
+      "step": 11232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4497829925031454,
+      "learning_rate": 2.6146499268681434e-05,
+      "loss": 1.125,
+      "step": 11233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.758288931207987,
+      "learning_rate": 2.6142333089680547e-05,
+      "loss": 0.8901,
+      "step": 11234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.413186531583367,
+      "learning_rate": 2.6138166878889003e-05,
+      "loss": 0.7979,
+      "step": 11235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.64936762768007,
+      "learning_rate": 2.6134000636422736e-05,
+      "loss": 0.834,
+      "step": 11236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5896828898688105,
+      "learning_rate": 2.6129834362397688e-05,
+      "loss": 0.8896,
+      "step": 11237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5213079339550615,
+      "learning_rate": 2.6125668056929815e-05,
+      "loss": 1.0439,
+      "step": 11238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.343856809817518,
+      "learning_rate": 2.6121501720135055e-05,
+      "loss": 1.0488,
+      "step": 11239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.638229008464431,
+      "learning_rate": 2.6117335352129364e-05,
+      "loss": 0.8936,
+      "step": 11240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.583289868697791,
+      "learning_rate": 2.6113168953028684e-05,
+      "loss": 1.1191,
+      "step": 11241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.906521637965781,
+      "learning_rate": 2.6109002522948962e-05,
+      "loss": 0.551,
+      "step": 11242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.188844285122097,
+      "learning_rate": 2.6104836062006145e-05,
+      "loss": 1.375,
+      "step": 11243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.013064865545337,
+      "learning_rate": 2.6100669570316195e-05,
+      "loss": 0.8516,
+      "step": 11244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.475729730721067,
+      "learning_rate": 2.6096503047995063e-05,
+      "loss": 1.4678,
+      "step": 11245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9371246747423694,
+      "learning_rate": 2.60923364951587e-05,
+      "loss": 0.5806,
+      "step": 11246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.404645916095348,
+      "learning_rate": 2.608816991192305e-05,
+      "loss": 0.7827,
+      "step": 11247
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1059453250668607,
+      "learning_rate": 2.6084003298404074e-05,
+      "loss": 1.1445,
+      "step": 11248
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.691701154851053,
+      "learning_rate": 2.607983665471773e-05,
+      "loss": 0.7285,
+      "step": 11249
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.946675912654767,
+      "learning_rate": 2.6075669980979976e-05,
+      "loss": 1.3931,
+      "step": 11250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7382666224199177,
+      "learning_rate": 2.6071503277306753e-05,
+      "loss": 1.4648,
+      "step": 11251
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9024661463193304,
+      "learning_rate": 2.6067336543814043e-05,
+      "loss": 1.2227,
+      "step": 11252
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.285953746978602,
+      "learning_rate": 2.6063169780617778e-05,
+      "loss": 1.5332,
+      "step": 11253
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.284371109109845,
+      "learning_rate": 2.6059002987833948e-05,
+      "loss": 0.6665,
+      "step": 11254
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.483986097835675,
+      "learning_rate": 2.6054836165578483e-05,
+      "loss": 1.4141,
+      "step": 11255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.435922156521278,
+      "learning_rate": 2.6050669313967358e-05,
+      "loss": 1.2598,
+      "step": 11256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.274112895764504,
+      "learning_rate": 2.6046502433116536e-05,
+      "loss": 1.0776,
+      "step": 11257
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.087172353466515,
+      "learning_rate": 2.604233552314198e-05,
+      "loss": 0.926,
+      "step": 11258
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.080229818939886,
+      "learning_rate": 2.6038168584159643e-05,
+      "loss": 0.8428,
+      "step": 11259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.807441983297547,
+      "learning_rate": 2.6034001616285507e-05,
+      "loss": 0.6641,
+      "step": 11260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.214048488884978,
+      "learning_rate": 2.6029834619635523e-05,
+      "loss": 1.3086,
+      "step": 11261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.634543975820029,
+      "learning_rate": 2.602566759432566e-05,
+      "loss": 1.3867,
+      "step": 11262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.35136243475288,
+      "learning_rate": 2.6021500540471884e-05,
+      "loss": 1.8447,
+      "step": 11263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.187793494017898,
+      "learning_rate": 2.6017333458190163e-05,
+      "loss": 0.877,
+      "step": 11264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.728165447722354,
+      "learning_rate": 2.6013166347596467e-05,
+      "loss": 1.1914,
+      "step": 11265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.716377808564103,
+      "learning_rate": 2.6008999208806762e-05,
+      "loss": 0.9191,
+      "step": 11266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.514256887188663,
+      "learning_rate": 2.6004832041937028e-05,
+      "loss": 1.1641,
+      "step": 11267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.300066664051291,
+      "learning_rate": 2.6000664847103222e-05,
+      "loss": 1.2227,
+      "step": 11268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.549723056639262,
+      "learning_rate": 2.5996497624421324e-05,
+      "loss": 1.1045,
+      "step": 11269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2840092315286844,
+      "learning_rate": 2.599233037400729e-05,
+      "loss": 0.5125,
+      "step": 11270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7398722364939645,
+      "learning_rate": 2.598816309597712e-05,
+      "loss": 0.6216,
+      "step": 11271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7662014980002927,
+      "learning_rate": 2.598399579044677e-05,
+      "loss": 1.209,
+      "step": 11272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.464133376394504,
+      "learning_rate": 2.5979828457532217e-05,
+      "loss": 0.9619,
+      "step": 11273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9771386737809786,
+      "learning_rate": 2.5975661097349428e-05,
+      "loss": 0.9779,
+      "step": 11274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1899687238732533,
+      "learning_rate": 2.59714937100144e-05,
+      "loss": 1.3789,
+      "step": 11275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.133419748562058,
+      "learning_rate": 2.596732629564309e-05,
+      "loss": 1.5518,
+      "step": 11276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.2953184648369325,
+      "learning_rate": 2.5963158854351492e-05,
+      "loss": 1.584,
+      "step": 11277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.014720799593531,
+      "learning_rate": 2.595899138625557e-05,
+      "loss": 0.7715,
+      "step": 11278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.229688675695901,
+      "learning_rate": 2.59548238914713e-05,
+      "loss": 1.1553,
+      "step": 11279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0768053267320994,
+      "learning_rate": 2.5950656370114684e-05,
+      "loss": 1.2168,
+      "step": 11280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.523654152933008,
+      "learning_rate": 2.5946488822301686e-05,
+      "loss": 1.2148,
+      "step": 11281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3463433731513903,
+      "learning_rate": 2.5942321248148288e-05,
+      "loss": 1.1953,
+      "step": 11282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4914041561621514,
+      "learning_rate": 2.593815364777047e-05,
+      "loss": 1.0527,
+      "step": 11283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.020432999536296,
+      "learning_rate": 2.593398602128421e-05,
+      "loss": 0.8748,
+      "step": 11284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6345031085933206,
+      "learning_rate": 2.5929818368805507e-05,
+      "loss": 0.9033,
+      "step": 11285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.211017037128303,
+      "learning_rate": 2.5925650690450343e-05,
+      "loss": 1.0415,
+      "step": 11286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046087448384244,
+      "learning_rate": 2.5921482986334688e-05,
+      "loss": 0.9346,
+      "step": 11287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.952238851995812,
+      "learning_rate": 2.591731525657454e-05,
+      "loss": 0.8594,
+      "step": 11288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.294930880548409,
+      "learning_rate": 2.5913147501285885e-05,
+      "loss": 1.4844,
+      "step": 11289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.678889386515301,
+      "learning_rate": 2.590897972058471e-05,
+      "loss": 0.8965,
+      "step": 11290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9512164695091996,
+      "learning_rate": 2.590481191458699e-05,
+      "loss": 0.9102,
+      "step": 11291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.486728091471292,
+      "learning_rate": 2.590064408340873e-05,
+      "loss": 1.5664,
+      "step": 11292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.812694595167807,
+      "learning_rate": 2.58964762271659e-05,
+      "loss": 0.8516,
+      "step": 11293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.737374627713114,
+      "learning_rate": 2.5892308345974515e-05,
+      "loss": 1.2969,
+      "step": 11294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0642639477685814,
+      "learning_rate": 2.5888140439950542e-05,
+      "loss": 0.9082,
+      "step": 11295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.640796182988016,
+      "learning_rate": 2.5883972509209992e-05,
+      "loss": 0.8857,
+      "step": 11296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.487123804064032,
+      "learning_rate": 2.587980455386884e-05,
+      "loss": 0.5536,
+      "step": 11297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.216659958430109,
+      "learning_rate": 2.5875636574043087e-05,
+      "loss": 0.4399,
+      "step": 11298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.524613482339608,
+      "learning_rate": 2.5871468569848727e-05,
+      "loss": 0.5059,
+      "step": 11299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.979110788385947,
+      "learning_rate": 2.586730054140175e-05,
+      "loss": 0.918,
+      "step": 11300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5237561946623646,
+      "learning_rate": 2.5863132488818154e-05,
+      "loss": 1.2129,
+      "step": 11301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2333840260042797,
+      "learning_rate": 2.585896441221392e-05,
+      "loss": 1.3691,
+      "step": 11302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.24784079898886,
+      "learning_rate": 2.585479631170507e-05,
+      "loss": 1.2168,
+      "step": 11303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9211937542080146,
+      "learning_rate": 2.5850628187407584e-05,
+      "loss": 1.0049,
+      "step": 11304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.418207652872451,
+      "learning_rate": 2.5846460039437466e-05,
+      "loss": 0.9355,
+      "step": 11305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.326507825958384,
+      "learning_rate": 2.58422918679107e-05,
+      "loss": 0.876,
+      "step": 11306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7613767765587367,
+      "learning_rate": 2.5838123672943298e-05,
+      "loss": 0.6016,
+      "step": 11307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2700065600934796,
+      "learning_rate": 2.5833955454651254e-05,
+      "loss": 0.9932,
+      "step": 11308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4970304018246794,
+      "learning_rate": 2.5829787213150575e-05,
+      "loss": 0.7949,
+      "step": 11309
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.567866683297242,
+      "learning_rate": 2.5825618948557244e-05,
+      "loss": 1.1201,
+      "step": 11310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.823198509704377,
+      "learning_rate": 2.582145066098729e-05,
+      "loss": 1.7559,
+      "step": 11311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.23933809105717,
+      "learning_rate": 2.5817282350556683e-05,
+      "loss": 0.7422,
+      "step": 11312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.222576751101786,
+      "learning_rate": 2.5813114017381452e-05,
+      "loss": 0.8164,
+      "step": 11313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6164032173765643,
+      "learning_rate": 2.5808945661577583e-05,
+      "loss": 1.0088,
+      "step": 11314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.433586500227464,
+      "learning_rate": 2.580477728326109e-05,
+      "loss": 1.207,
+      "step": 11315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0169167395644774,
+      "learning_rate": 2.5800608882547965e-05,
+      "loss": 0.8711,
+      "step": 11316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.204817642490962,
+      "learning_rate": 2.5796440459554228e-05,
+      "loss": 0.5567,
+      "step": 11317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.064608099670405,
+      "learning_rate": 2.5792272014395874e-05,
+      "loss": 1.4062,
+      "step": 11318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.624425007651881,
+      "learning_rate": 2.5788103547188915e-05,
+      "loss": 0.9385,
+      "step": 11319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.253365436904615,
+      "learning_rate": 2.5783935058049358e-05,
+      "loss": 0.481,
+      "step": 11320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.981391537723985,
+      "learning_rate": 2.5779766547093194e-05,
+      "loss": 1.4141,
+      "step": 11321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.374110104110419,
+      "learning_rate": 2.577559801443646e-05,
+      "loss": 1.4082,
+      "step": 11322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.633578705741136,
+      "learning_rate": 2.5771429460195136e-05,
+      "loss": 0.8145,
+      "step": 11323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.921678521003919,
+      "learning_rate": 2.5767260884485255e-05,
+      "loss": 1.3848,
+      "step": 11324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.353724922487366,
+      "learning_rate": 2.57630922874228e-05,
+      "loss": 1.0054,
+      "step": 11325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7669809756546955,
+      "learning_rate": 2.575892366912382e-05,
+      "loss": 1.3047,
+      "step": 11326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.089103332945363,
+      "learning_rate": 2.5754755029704286e-05,
+      "loss": 1.4746,
+      "step": 11327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.757457946420187,
+      "learning_rate": 2.5750586369280238e-05,
+      "loss": 0.9365,
+      "step": 11328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4014948639934897,
+      "learning_rate": 2.5746417687967672e-05,
+      "loss": 1.5566,
+      "step": 11329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0198018171328505,
+      "learning_rate": 2.57422489858826e-05,
+      "loss": 0.9111,
+      "step": 11330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8637608312554335,
+      "learning_rate": 2.573808026314105e-05,
+      "loss": 1.0178,
+      "step": 11331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.240016568930375,
+      "learning_rate": 2.5733911519859022e-05,
+      "loss": 1.5815,
+      "step": 11332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3243666368305074,
+      "learning_rate": 2.5729742756152536e-05,
+      "loss": 0.8579,
+      "step": 11333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.92800528526545,
+      "learning_rate": 2.5725573972137605e-05,
+      "loss": 1.3447,
+      "step": 11334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.114727209548996,
+      "learning_rate": 2.5721405167930253e-05,
+      "loss": 0.803,
+      "step": 11335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.188058388901432,
+      "learning_rate": 2.5717236343646478e-05,
+      "loss": 1.1895,
+      "step": 11336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.779235016329349,
+      "learning_rate": 2.5713067499402317e-05,
+      "loss": 0.7539,
+      "step": 11337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8425204163445135,
+      "learning_rate": 2.570889863531377e-05,
+      "loss": 0.8184,
+      "step": 11338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.868291807617821,
+      "learning_rate": 2.5704729751496864e-05,
+      "loss": 0.4028,
+      "step": 11339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3268245251596325,
+      "learning_rate": 2.5700560848067618e-05,
+      "loss": 0.5713,
+      "step": 11340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.085044398539275,
+      "learning_rate": 2.569639192514205e-05,
+      "loss": 0.5928,
+      "step": 11341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8188372612955472,
+      "learning_rate": 2.569222298283618e-05,
+      "loss": 1.3213,
+      "step": 11342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.085966675400737,
+      "learning_rate": 2.5688054021266022e-05,
+      "loss": 0.8076,
+      "step": 11343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9448154534504647,
+      "learning_rate": 2.5683885040547595e-05,
+      "loss": 0.4535,
+      "step": 11344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.027176054652617,
+      "learning_rate": 2.5679716040796937e-05,
+      "loss": 0.9277,
+      "step": 11345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.079894495635794,
+      "learning_rate": 2.5675547022130054e-05,
+      "loss": 0.8398,
+      "step": 11346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.003353771038664,
+      "learning_rate": 2.567137798466297e-05,
+      "loss": 0.8789,
+      "step": 11347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.491903354626015,
+      "learning_rate": 2.566720892851171e-05,
+      "loss": 1.1514,
+      "step": 11348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.493973584604227,
+      "learning_rate": 2.56630398537923e-05,
+      "loss": 0.8428,
+      "step": 11349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9753760028092686,
+      "learning_rate": 2.565887076062076e-05,
+      "loss": 1.4863,
+      "step": 11350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.229536138795592,
+      "learning_rate": 2.5654701649113118e-05,
+      "loss": 1.1465,
+      "step": 11351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4949725692174636,
+      "learning_rate": 2.565053251938539e-05,
+      "loss": 0.9136,
+      "step": 11352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.3148128691407495,
+      "learning_rate": 2.5646363371553605e-05,
+      "loss": 1.2354,
+      "step": 11353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9909174442306368,
+      "learning_rate": 2.56421942057338e-05,
+      "loss": 1.4199,
+      "step": 11354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6861593104525654,
+      "learning_rate": 2.5638025022041985e-05,
+      "loss": 0.6152,
+      "step": 11355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0040608578344816,
+      "learning_rate": 2.56338558205942e-05,
+      "loss": 0.9824,
+      "step": 11356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9519480096193853,
+      "learning_rate": 2.562968660150646e-05,
+      "loss": 0.7051,
+      "step": 11357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.018708110490153,
+      "learning_rate": 2.5625517364894798e-05,
+      "loss": 1.3516,
+      "step": 11358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.602000798142448,
+      "learning_rate": 2.5621348110875237e-05,
+      "loss": 1.3828,
+      "step": 11359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.281988968635288,
+      "learning_rate": 2.561717883956382e-05,
+      "loss": 1.2617,
+      "step": 11360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1370495639492395,
+      "learning_rate": 2.5613009551076567e-05,
+      "loss": 0.6182,
+      "step": 11361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9460576174838584,
+      "learning_rate": 2.560884024552951e-05,
+      "loss": 1.2168,
+      "step": 11362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.067616140192468,
+      "learning_rate": 2.560467092303866e-05,
+      "loss": 1.1445,
+      "step": 11363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.300187599170706,
+      "learning_rate": 2.5600501583720083e-05,
+      "loss": 1.1074,
+      "step": 11364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.260901135531329,
+      "learning_rate": 2.559633222768978e-05,
+      "loss": 1.3271,
+      "step": 11365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3784804691317136,
+      "learning_rate": 2.5592162855063807e-05,
+      "loss": 1.1387,
+      "step": 11366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5768498822009733,
+      "learning_rate": 2.558799346595817e-05,
+      "loss": 1.4883,
+      "step": 11367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.58428392810986,
+      "learning_rate": 2.5583824060488916e-05,
+      "loss": 1.7285,
+      "step": 11368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.514965563286182,
+      "learning_rate": 2.557965463877208e-05,
+      "loss": 0.959,
+      "step": 11369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.38744093714823,
+      "learning_rate": 2.5575485200923688e-05,
+      "loss": 0.8318,
+      "step": 11370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6442876245775957,
+      "learning_rate": 2.5571315747059788e-05,
+      "loss": 0.998,
+      "step": 11371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.792190642936628,
+      "learning_rate": 2.5567146277296385e-05,
+      "loss": 1.1377,
+      "step": 11372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.990641468831881,
+      "learning_rate": 2.5562976791749543e-05,
+      "loss": 1.4541,
+      "step": 11373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.136403285420457,
+      "learning_rate": 2.5558807290535285e-05,
+      "loss": 1.2812,
+      "step": 11374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3915430437984124,
+      "learning_rate": 2.555463777376965e-05,
+      "loss": 0.7832,
+      "step": 11375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.521345452955506,
+      "learning_rate": 2.5550468241568658e-05,
+      "loss": 0.7343,
+      "step": 11376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.829215498115658,
+      "learning_rate": 2.5546298694048376e-05,
+      "loss": 0.6477,
+      "step": 11377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.284846342258422,
+      "learning_rate": 2.5542129131324815e-05,
+      "loss": 1.0664,
+      "step": 11378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.486442849904086,
+      "learning_rate": 2.5537959553514028e-05,
+      "loss": 1.1484,
+      "step": 11379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.575442445492624,
+      "learning_rate": 2.553378996073204e-05,
+      "loss": 1.0537,
+      "step": 11380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.001370623601496,
+      "learning_rate": 2.5529620353094895e-05,
+      "loss": 1.3535,
+      "step": 11381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.070852825952418,
+      "learning_rate": 2.552545073071863e-05,
+      "loss": 1.4268,
+      "step": 11382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0281244268111718,
+      "learning_rate": 2.552128109371929e-05,
+      "loss": 1.0654,
+      "step": 11383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.557953057676889,
+      "learning_rate": 2.5517111442212903e-05,
+      "loss": 1.1846,
+      "step": 11384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.017121626122069,
+      "learning_rate": 2.5512941776315525e-05,
+      "loss": 0.9871,
+      "step": 11385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.809616305943054,
+      "learning_rate": 2.5508772096143175e-05,
+      "loss": 0.752,
+      "step": 11386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7794843920073733,
+      "learning_rate": 2.550460240181191e-05,
+      "loss": 0.8867,
+      "step": 11387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.240590882801606,
+      "learning_rate": 2.5500432693437764e-05,
+      "loss": 1.168,
+      "step": 11388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.473291249686784,
+      "learning_rate": 2.549626297113678e-05,
+      "loss": 1.3594,
+      "step": 11389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1343969251810715,
+      "learning_rate": 2.5492093235025e-05,
+      "loss": 0.9355,
+      "step": 11390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.848613804804829,
+      "learning_rate": 2.5487923485218463e-05,
+      "loss": 1.3369,
+      "step": 11391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.239882503902073,
+      "learning_rate": 2.548375372183322e-05,
+      "loss": 0.8818,
+      "step": 11392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.254287352361714,
+      "learning_rate": 2.5479583944985307e-05,
+      "loss": 0.8052,
+      "step": 11393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.915236874030796,
+      "learning_rate": 2.5475414154790767e-05,
+      "loss": 0.3921,
+      "step": 11394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.152926559696892,
+      "learning_rate": 2.5471244351365636e-05,
+      "loss": 1.251,
+      "step": 11395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.240503584628259,
+      "learning_rate": 2.546707453482598e-05,
+      "loss": 0.7402,
+      "step": 11396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7476215449211945,
+      "learning_rate": 2.5462904705287822e-05,
+      "loss": 1.1475,
+      "step": 11397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.821732013243892,
+      "learning_rate": 2.545873486286722e-05,
+      "loss": 0.9326,
+      "step": 11398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3191359905590483,
+      "learning_rate": 2.545456500768021e-05,
+      "loss": 0.874,
+      "step": 11399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.05274977070745,
+      "learning_rate": 2.5450395139842835e-05,
+      "loss": 1.7617,
+      "step": 11400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2337865616685684,
+      "learning_rate": 2.544622525947115e-05,
+      "loss": 1.0234,
+      "step": 11401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6105306658825955,
+      "learning_rate": 2.5442055366681205e-05,
+      "loss": 1.4961,
+      "step": 11402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.62974428727196,
+      "learning_rate": 2.5437885461589034e-05,
+      "loss": 0.7275,
+      "step": 11403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.252961029541356,
+      "learning_rate": 2.5433715544310688e-05,
+      "loss": 1.8047,
+      "step": 11404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5730582756650517,
+      "learning_rate": 2.5429545614962217e-05,
+      "loss": 0.957,
+      "step": 11405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8734919697928283,
+      "learning_rate": 2.5425375673659662e-05,
+      "loss": 0.9785,
+      "step": 11406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6319192791767096,
+      "learning_rate": 2.542120572051908e-05,
+      "loss": 1.2344,
+      "step": 11407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8924116232253225,
+      "learning_rate": 2.5417035755656517e-05,
+      "loss": 1.3157,
+      "step": 11408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.608616032589376,
+      "learning_rate": 2.541286577918801e-05,
+      "loss": 1.2305,
+      "step": 11409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1158814717815257,
+      "learning_rate": 2.540869579122962e-05,
+      "loss": 1.0249,
+      "step": 11410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.769211517039551,
+      "learning_rate": 2.54045257918974e-05,
+      "loss": 0.9854,
+      "step": 11411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8490436659629472,
+      "learning_rate": 2.5400355781307384e-05,
+      "loss": 0.4528,
+      "step": 11412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3685966782367167,
+      "learning_rate": 2.539618575957563e-05,
+      "loss": 0.6174,
+      "step": 11413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.475545557960709,
+      "learning_rate": 2.539201572681818e-05,
+      "loss": 1.4785,
+      "step": 11414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.81950641241675,
+      "learning_rate": 2.5387845683151107e-05,
+      "loss": 1.3379,
+      "step": 11415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0587466217550645,
+      "learning_rate": 2.538367562869044e-05,
+      "loss": 1.7012,
+      "step": 11416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.123240920869942,
+      "learning_rate": 2.5379505563552236e-05,
+      "loss": 0.8384,
+      "step": 11417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.253698045653864,
+      "learning_rate": 2.5375335487852546e-05,
+      "loss": 1.1875,
+      "step": 11418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.59141808224977,
+      "learning_rate": 2.5371165401707425e-05,
+      "loss": 0.8818,
+      "step": 11419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3993408291314426,
+      "learning_rate": 2.5366995305232916e-05,
+      "loss": 0.8589,
+      "step": 11420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7846037763121916,
+      "learning_rate": 2.5362825198545088e-05,
+      "loss": 1.0215,
+      "step": 11421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.410157060761692,
+      "learning_rate": 2.535865508175997e-05,
+      "loss": 1.5469,
+      "step": 11422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.604455204555575,
+      "learning_rate": 2.5354484954993634e-05,
+      "loss": 1.3203,
+      "step": 11423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.08214642667954,
+      "learning_rate": 2.5350314818362124e-05,
+      "loss": 1.5684,
+      "step": 11424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2664113535335937,
+      "learning_rate": 2.5346144671981498e-05,
+      "loss": 0.959,
+      "step": 11425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7394013338348513,
+      "learning_rate": 2.534197451596781e-05,
+      "loss": 1.5352,
+      "step": 11426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8782101172329784,
+      "learning_rate": 2.5337804350437094e-05,
+      "loss": 1.1445,
+      "step": 11427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.663127030159446,
+      "learning_rate": 2.5333634175505437e-05,
+      "loss": 0.6357,
+      "step": 11428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.661635739459912,
+      "learning_rate": 2.532946399128887e-05,
+      "loss": 0.9351,
+      "step": 11429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.657398200108371,
+      "learning_rate": 2.532529379790346e-05,
+      "loss": 1.5703,
+      "step": 11430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.474617883544248,
+      "learning_rate": 2.5321123595465246e-05,
+      "loss": 0.9053,
+      "step": 11431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.800978902499073,
+      "learning_rate": 2.5316953384090298e-05,
+      "loss": 1.0918,
+      "step": 11432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.109098747787954,
+      "learning_rate": 2.5312783163894664e-05,
+      "loss": 0.7656,
+      "step": 11433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.817601652159737,
+      "learning_rate": 2.5308612934994415e-05,
+      "loss": 1.4453,
+      "step": 11434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.062338286140544,
+      "learning_rate": 2.5304442697505577e-05,
+      "loss": 0.8882,
+      "step": 11435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.135157045874137,
+      "learning_rate": 2.5300272451544234e-05,
+      "loss": 1.2207,
+      "step": 11436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3066223298848447,
+      "learning_rate": 2.5296102197226423e-05,
+      "loss": 0.7842,
+      "step": 11437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.005019852759296,
+      "learning_rate": 2.529193193466821e-05,
+      "loss": 0.9531,
+      "step": 11438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7037699376542634,
+      "learning_rate": 2.5287761663985653e-05,
+      "loss": 1.5117,
+      "step": 11439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.8240910291579375,
+      "learning_rate": 2.5283591385294808e-05,
+      "loss": 1.0503,
+      "step": 11440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.512347153147038,
+      "learning_rate": 2.5279421098711727e-05,
+      "loss": 1.1582,
+      "step": 11441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6842330426410044,
+      "learning_rate": 2.5275250804352463e-05,
+      "loss": 0.8672,
+      "step": 11442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.514731775773164,
+      "learning_rate": 2.5271080502333094e-05,
+      "loss": 1.1445,
+      "step": 11443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.223126001511367,
+      "learning_rate": 2.526691019276966e-05,
+      "loss": 1.3477,
+      "step": 11444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2947302945300105,
+      "learning_rate": 2.5262739875778225e-05,
+      "loss": 1.1582,
+      "step": 11445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.898579621268619,
+      "learning_rate": 2.5258569551474838e-05,
+      "loss": 1.085,
+      "step": 11446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.522721132806778,
+      "learning_rate": 2.525439921997558e-05,
+      "loss": 1.293,
+      "step": 11447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.004925657803091,
+      "learning_rate": 2.5250228881396482e-05,
+      "loss": 0.5537,
+      "step": 11448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1326457087655726,
+      "learning_rate": 2.524605853585363e-05,
+      "loss": 0.6279,
+      "step": 11449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.875948486901097,
+      "learning_rate": 2.5241888183463054e-05,
+      "loss": 2.125,
+      "step": 11450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.109816554443087,
+      "learning_rate": 2.5237717824340833e-05,
+      "loss": 0.8647,
+      "step": 11451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5701251321480805,
+      "learning_rate": 2.5233547458603025e-05,
+      "loss": 0.8301,
+      "step": 11452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.040287267461719,
+      "learning_rate": 2.5229377086365685e-05,
+      "loss": 0.7734,
+      "step": 11453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2693069531012764,
+      "learning_rate": 2.5225206707744874e-05,
+      "loss": 0.9863,
+      "step": 11454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.006366808360812,
+      "learning_rate": 2.522103632285665e-05,
+      "loss": 1.2217,
+      "step": 11455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.267055173908316,
+      "learning_rate": 2.5216865931817086e-05,
+      "loss": 1.498,
+      "step": 11456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.968162365358332,
+      "learning_rate": 2.521269553474222e-05,
+      "loss": 0.7263,
+      "step": 11457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.784371927689935,
+      "learning_rate": 2.5208525131748127e-05,
+      "loss": 1.1404,
+      "step": 11458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3339570574543558,
+      "learning_rate": 2.5204354722950863e-05,
+      "loss": 1.7969,
+      "step": 11459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2207009777639137,
+      "learning_rate": 2.520018430846649e-05,
+      "loss": 0.9351,
+      "step": 11460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.257959429804302,
+      "learning_rate": 2.519601388841107e-05,
+      "loss": 1.3242,
+      "step": 11461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9795954658052475,
+      "learning_rate": 2.5191843462900666e-05,
+      "loss": 0.4438,
+      "step": 11462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6895577293337554,
+      "learning_rate": 2.5187673032051336e-05,
+      "loss": 0.8408,
+      "step": 11463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3581293114431414,
+      "learning_rate": 2.5183502595979137e-05,
+      "loss": 0.9512,
+      "step": 11464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.823092202972977,
+      "learning_rate": 2.517933215480014e-05,
+      "loss": 1.7734,
+      "step": 11465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4574980421242816,
+      "learning_rate": 2.5175161708630402e-05,
+      "loss": 1.6504,
+      "step": 11466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.29465673948074,
+      "learning_rate": 2.517099125758598e-05,
+      "loss": 0.5339,
+      "step": 11467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6472550299560953,
+      "learning_rate": 2.516682080178295e-05,
+      "loss": 0.9238,
+      "step": 11468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.908731267607878,
+      "learning_rate": 2.5162650341337353e-05,
+      "loss": 1.3398,
+      "step": 11469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2839216708879255,
+      "learning_rate": 2.5158479876365277e-05,
+      "loss": 1.5703,
+      "step": 11470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8648563641584626,
+      "learning_rate": 2.5154309406982755e-05,
+      "loss": 0.8896,
+      "step": 11471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.042157933112657,
+      "learning_rate": 2.5150138933305884e-05,
+      "loss": 1.124,
+      "step": 11472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9525120013168316,
+      "learning_rate": 2.5145968455450692e-05,
+      "loss": 0.9453,
+      "step": 11473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.208073659189141,
+      "learning_rate": 2.514179797353326e-05,
+      "loss": 1.0859,
+      "step": 11474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.286366929390234,
+      "learning_rate": 2.5137627487669657e-05,
+      "loss": 1.3125,
+      "step": 11475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.442097167860138,
+      "learning_rate": 2.5133456997975928e-05,
+      "loss": 0.8574,
+      "step": 11476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.004321958749383,
+      "learning_rate": 2.512928650456815e-05,
+      "loss": 0.7083,
+      "step": 11477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.340628618102594,
+      "learning_rate": 2.5125116007562373e-05,
+      "loss": 0.7893,
+      "step": 11478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.960195795317041,
+      "learning_rate": 2.5120945507074676e-05,
+      "loss": 1.0293,
+      "step": 11479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.297905395107381,
+      "learning_rate": 2.5116775003221115e-05,
+      "loss": 0.9487,
+      "step": 11480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.333731071266815,
+      "learning_rate": 2.5112604496117753e-05,
+      "loss": 1.1553,
+      "step": 11481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8851765713345374,
+      "learning_rate": 2.5108433985880653e-05,
+      "loss": 0.8047,
+      "step": 11482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.006150774993601,
+      "learning_rate": 2.5104263472625877e-05,
+      "loss": 1.041,
+      "step": 11483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.063705051548419,
+      "learning_rate": 2.510009295646949e-05,
+      "loss": 1.0664,
+      "step": 11484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.904078497988614,
+      "learning_rate": 2.5095922437527562e-05,
+      "loss": 1.4297,
+      "step": 11485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.286962109515078,
+      "learning_rate": 2.5091751915916145e-05,
+      "loss": 0.9131,
+      "step": 11486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7249910962524615,
+      "learning_rate": 2.508758139175132e-05,
+      "loss": 1.2441,
+      "step": 11487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.09180889639766,
+      "learning_rate": 2.5083410865149127e-05,
+      "loss": 1.8906,
+      "step": 11488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5628506490330776,
+      "learning_rate": 2.507924033622565e-05,
+      "loss": 1.0137,
+      "step": 11489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.024715943160194,
+      "learning_rate": 2.5075069805096944e-05,
+      "loss": 1.0039,
+      "step": 11490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.123774979014865,
+      "learning_rate": 2.507089927187909e-05,
+      "loss": 1.0923,
+      "step": 11491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2784552888693974,
+      "learning_rate": 2.506672873668812e-05,
+      "loss": 1.3145,
+      "step": 11492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.47195385228197,
+      "learning_rate": 2.5062558199640124e-05,
+      "loss": 0.709,
+      "step": 11493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7779628873489526,
+      "learning_rate": 2.5058387660851163e-05,
+      "loss": 0.8083,
+      "step": 11494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.563646478743677,
+      "learning_rate": 2.505421712043729e-05,
+      "loss": 1.0293,
+      "step": 11495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.852677343634036,
+      "learning_rate": 2.5050046578514587e-05,
+      "loss": 0.8525,
+      "step": 11496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.468300034357765,
+      "learning_rate": 2.504587603519909e-05,
+      "loss": 1.457,
+      "step": 11497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.029355089165512,
+      "learning_rate": 2.5041705490606892e-05,
+      "loss": 1.2793,
+      "step": 11498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.221062372639255,
+      "learning_rate": 2.5037534944854046e-05,
+      "loss": 0.958,
+      "step": 11499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.457968129444136,
+      "learning_rate": 2.5033364398056623e-05,
+      "loss": 1.3691,
+      "step": 11500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.349134338506494,
+      "learning_rate": 2.502919385033068e-05,
+      "loss": 1.04,
+      "step": 11501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.866629432291923,
+      "learning_rate": 2.5025023301792277e-05,
+      "loss": 1.0078,
+      "step": 11502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9845366800453053,
+      "learning_rate": 2.502085275255749e-05,
+      "loss": 1.0537,
+      "step": 11503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.783560599711098,
+      "learning_rate": 2.5016682202742386e-05,
+      "loss": 0.8555,
+      "step": 11504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.387929931027025,
+      "learning_rate": 2.5012511652463016e-05,
+      "loss": 1.3594,
+      "step": 11505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.749254819303297,
+      "learning_rate": 2.500834110183545e-05,
+      "loss": 1.1282,
+      "step": 11506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.7017211866813495,
+      "learning_rate": 2.5004170550975764e-05,
+      "loss": 1.8145,
+      "step": 11507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4711247864618477,
+      "learning_rate": 2.5e-05,
+      "loss": 0.8125,
+      "step": 11508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.507697842552698,
+      "learning_rate": 2.4995829449024246e-05,
+      "loss": 1.0762,
+      "step": 11509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0445167992867943,
+      "learning_rate": 2.4991658898164554e-05,
+      "loss": 0.4983,
+      "step": 11510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6914647155523843,
+      "learning_rate": 2.4987488347536993e-05,
+      "loss": 1.084,
+      "step": 11511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.257260648610025,
+      "learning_rate": 2.4983317797257627e-05,
+      "loss": 1.084,
+      "step": 11512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.194199171045811,
+      "learning_rate": 2.497914724744251e-05,
+      "loss": 0.8643,
+      "step": 11513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.817951006261064,
+      "learning_rate": 2.4974976698207722e-05,
+      "loss": 1.3203,
+      "step": 11514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.663023894440232,
+      "learning_rate": 2.4970806149669327e-05,
+      "loss": 0.6824,
+      "step": 11515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0186162946641,
+      "learning_rate": 2.496663560194338e-05,
+      "loss": 1.6152,
+      "step": 11516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.651736126094902,
+      "learning_rate": 2.4962465055145956e-05,
+      "loss": 0.8794,
+      "step": 11517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.742089360084164,
+      "learning_rate": 2.495829450939311e-05,
+      "loss": 1.2236,
+      "step": 11518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7060473536241663,
+      "learning_rate": 2.4954123964800917e-05,
+      "loss": 0.8438,
+      "step": 11519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4003508947078154,
+      "learning_rate": 2.494995342148543e-05,
+      "loss": 1.4375,
+      "step": 11520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.128845101974468,
+      "learning_rate": 2.494578287956272e-05,
+      "loss": 0.9736,
+      "step": 11521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1615281492733005,
+      "learning_rate": 2.4941612339148842e-05,
+      "loss": 0.7051,
+      "step": 11522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.849607425773176,
+      "learning_rate": 2.4937441800359875e-05,
+      "loss": 1.2969,
+      "step": 11523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0640265855111997,
+      "learning_rate": 2.4933271263311882e-05,
+      "loss": 0.9775,
+      "step": 11524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.984555393617586,
+      "learning_rate": 2.4929100728120917e-05,
+      "loss": 1.0857,
+      "step": 11525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.874822550986786,
+      "learning_rate": 2.4924930194903058e-05,
+      "loss": 1.1191,
+      "step": 11526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4081237235364146,
+      "learning_rate": 2.4920759663774355e-05,
+      "loss": 0.6313,
+      "step": 11527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4268083102769165,
+      "learning_rate": 2.491658913485088e-05,
+      "loss": 0.9497,
+      "step": 11528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0105589378731987,
+      "learning_rate": 2.4912418608248693e-05,
+      "loss": 1.083,
+      "step": 11529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1601513936710783,
+      "learning_rate": 2.490824808408386e-05,
+      "loss": 0.6085,
+      "step": 11530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9433882116412557,
+      "learning_rate": 2.490407756247245e-05,
+      "loss": 0.7942,
+      "step": 11531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.815634267734829,
+      "learning_rate": 2.489990704353051e-05,
+      "loss": 0.7344,
+      "step": 11532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4009439996037805,
+      "learning_rate": 2.4895736527374126e-05,
+      "loss": 0.6279,
+      "step": 11533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8086525431689806,
+      "learning_rate": 2.4891566014119356e-05,
+      "loss": 0.7563,
+      "step": 11534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8719016224491627,
+      "learning_rate": 2.4887395503882253e-05,
+      "loss": 1.3672,
+      "step": 11535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5057713648503634,
+      "learning_rate": 2.488322499677889e-05,
+      "loss": 0.6859,
+      "step": 11536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3726448629066454,
+      "learning_rate": 2.4879054492925326e-05,
+      "loss": 1.1235,
+      "step": 11537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.592392493803058,
+      "learning_rate": 2.4874883992437633e-05,
+      "loss": 1.0498,
+      "step": 11538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2247139736164505,
+      "learning_rate": 2.487071349543186e-05,
+      "loss": 1.2129,
+      "step": 11539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8472734134969775,
+      "learning_rate": 2.4866543002024084e-05,
+      "loss": 0.7043,
+      "step": 11540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.045423532905932,
+      "learning_rate": 2.486237251233035e-05,
+      "loss": 2.041,
+      "step": 11541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2017901549937036,
+      "learning_rate": 2.485820202646674e-05,
+      "loss": 0.644,
+      "step": 11542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3004742541685124,
+      "learning_rate": 2.4854031544549314e-05,
+      "loss": 1.1113,
+      "step": 11543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4045168096629435,
+      "learning_rate": 2.4849861066694125e-05,
+      "loss": 0.957,
+      "step": 11544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1727551110966536,
+      "learning_rate": 2.4845690593017247e-05,
+      "loss": 1.2783,
+      "step": 11545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.9552780418298275,
+      "learning_rate": 2.4841520123634732e-05,
+      "loss": 1.3892,
+      "step": 11546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4207427359388656,
+      "learning_rate": 2.4837349658662653e-05,
+      "loss": 0.75,
+      "step": 11547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.221467367135837,
+      "learning_rate": 2.483317919821706e-05,
+      "loss": 1.5918,
+      "step": 11548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.590087749503249,
+      "learning_rate": 2.482900874241402e-05,
+      "loss": 0.6543,
+      "step": 11549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6728585200973565,
+      "learning_rate": 2.4824838291369597e-05,
+      "loss": 1.4316,
+      "step": 11550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6716779075744117,
+      "learning_rate": 2.4820667845199864e-05,
+      "loss": 1.1592,
+      "step": 11551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.626754174622453,
+      "learning_rate": 2.4816497404020862e-05,
+      "loss": 1.4551,
+      "step": 11552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2883062057202257,
+      "learning_rate": 2.481232696794867e-05,
+      "loss": 1.6875,
+      "step": 11553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9929381541991207,
+      "learning_rate": 2.480815653709934e-05,
+      "loss": 0.7988,
+      "step": 11554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.544614947555571,
+      "learning_rate": 2.4803986111588937e-05,
+      "loss": 1.0801,
+      "step": 11555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.10214835142133,
+      "learning_rate": 2.4799815691533514e-05,
+      "loss": 0.5508,
+      "step": 11556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2774024811854874,
+      "learning_rate": 2.479564527704915e-05,
+      "loss": 0.7231,
+      "step": 11557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9205523266394593,
+      "learning_rate": 2.4791474868251882e-05,
+      "loss": 0.7095,
+      "step": 11558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.595277078304593,
+      "learning_rate": 2.4787304465257783e-05,
+      "loss": 1.127,
+      "step": 11559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9806540739189205,
+      "learning_rate": 2.4783134068182923e-05,
+      "loss": 1.3232,
+      "step": 11560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7670695524749,
+      "learning_rate": 2.477896367714335e-05,
+      "loss": 0.6802,
+      "step": 11561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.884846005300692,
+      "learning_rate": 2.477479329225513e-05,
+      "loss": 2.248,
+      "step": 11562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.12034367248293,
+      "learning_rate": 2.4770622913634318e-05,
+      "loss": 0.7939,
+      "step": 11563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.406889297969868,
+      "learning_rate": 2.476645254139698e-05,
+      "loss": 1.0195,
+      "step": 11564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.249718287069281,
+      "learning_rate": 2.4762282175659172e-05,
+      "loss": 0.9971,
+      "step": 11565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3974864812923045,
+      "learning_rate": 2.4758111816536955e-05,
+      "loss": 0.6743,
+      "step": 11566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0215476583648355,
+      "learning_rate": 2.4753941464146387e-05,
+      "loss": 1.0352,
+      "step": 11567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.485946734362778,
+      "learning_rate": 2.474977111860353e-05,
+      "loss": 1.8418,
+      "step": 11568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.572158206034732,
+      "learning_rate": 2.4745600780024424e-05,
+      "loss": 1.5527,
+      "step": 11569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.200204252638565,
+      "learning_rate": 2.4741430448525164e-05,
+      "loss": 0.7229,
+      "step": 11570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.284701333851347,
+      "learning_rate": 2.473726012422178e-05,
+      "loss": 1.0518,
+      "step": 11571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6696470017271134,
+      "learning_rate": 2.4733089807230346e-05,
+      "loss": 1.0146,
+      "step": 11572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4805441056467012,
+      "learning_rate": 2.4728919497666912e-05,
+      "loss": 0.532,
+      "step": 11573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8889373428111482,
+      "learning_rate": 2.472474919564754e-05,
+      "loss": 1.126,
+      "step": 11574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.121849385441317,
+      "learning_rate": 2.4720578901288282e-05,
+      "loss": 1.0938,
+      "step": 11575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2193597012428308,
+      "learning_rate": 2.4716408614705204e-05,
+      "loss": 1.4883,
+      "step": 11576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.151520227087614,
+      "learning_rate": 2.4712238336014353e-05,
+      "loss": 0.5503,
+      "step": 11577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.071479209506453,
+      "learning_rate": 2.4708068065331787e-05,
+      "loss": 0.9697,
+      "step": 11578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.836284521640118,
+      "learning_rate": 2.4703897802773583e-05,
+      "loss": 0.6606,
+      "step": 11579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.465516314241701,
+      "learning_rate": 2.4699727548455772e-05,
+      "loss": 1.0762,
+      "step": 11580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1832695517311547,
+      "learning_rate": 2.4695557302494426e-05,
+      "loss": 0.9619,
+      "step": 11581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.263573711067208,
+      "learning_rate": 2.4691387065005594e-05,
+      "loss": 0.6199,
+      "step": 11582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6349454214679646,
+      "learning_rate": 2.4687216836105338e-05,
+      "loss": 0.7148,
+      "step": 11583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8881847894655484,
+      "learning_rate": 2.4683046615909704e-05,
+      "loss": 0.6674,
+      "step": 11584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8025299529381447,
+      "learning_rate": 2.4678876404534763e-05,
+      "loss": 0.6826,
+      "step": 11585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9416124491967888,
+      "learning_rate": 2.4674706202096555e-05,
+      "loss": 1.1172,
+      "step": 11586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2429057708689126,
+      "learning_rate": 2.4670536008711133e-05,
+      "loss": 0.5757,
+      "step": 11587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0686446471341915,
+      "learning_rate": 2.4666365824494565e-05,
+      "loss": 1.5566,
+      "step": 11588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4840673473343937,
+      "learning_rate": 2.4662195649562905e-05,
+      "loss": 1.498,
+      "step": 11589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6493840102789923,
+      "learning_rate": 2.46580254840322e-05,
+      "loss": 0.9258,
+      "step": 11590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9240084616660247,
+      "learning_rate": 2.4653855328018508e-05,
+      "loss": 0.5098,
+      "step": 11591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3511597006732616,
+      "learning_rate": 2.464968518163788e-05,
+      "loss": 1.1836,
+      "step": 11592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1912462248071978,
+      "learning_rate": 2.4645515045006375e-05,
+      "loss": 1.123,
+      "step": 11593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6092127218438956,
+      "learning_rate": 2.4641344918240033e-05,
+      "loss": 1.1406,
+      "step": 11594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.099459071802814,
+      "learning_rate": 2.4637174801454925e-05,
+      "loss": 0.9937,
+      "step": 11595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.120095545207143,
+      "learning_rate": 2.463300469476709e-05,
+      "loss": 0.9224,
+      "step": 11596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.746941898044323,
+      "learning_rate": 2.4628834598292578e-05,
+      "loss": 0.7944,
+      "step": 11597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.389244367185734,
+      "learning_rate": 2.4624664512147456e-05,
+      "loss": 1.5059,
+      "step": 11598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.072678476898991,
+      "learning_rate": 2.4620494436447766e-05,
+      "loss": 1.668,
+      "step": 11599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.375235973767258,
+      "learning_rate": 2.4616324371309567e-05,
+      "loss": 0.6934,
+      "step": 11600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.839974627527197,
+      "learning_rate": 2.46121543168489e-05,
+      "loss": 0.4084,
+      "step": 11601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.434926443435701,
+      "learning_rate": 2.4607984273181822e-05,
+      "loss": 1.0049,
+      "step": 11602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.577246591448623,
+      "learning_rate": 2.460381424042438e-05,
+      "loss": 0.5508,
+      "step": 11603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.10194626864586,
+      "learning_rate": 2.459964421869263e-05,
+      "loss": 1.123,
+      "step": 11604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.797262468128084,
+      "learning_rate": 2.4595474208102616e-05,
+      "loss": 1.2715,
+      "step": 11605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7609656789622927,
+      "learning_rate": 2.459130420877038e-05,
+      "loss": 0.7295,
+      "step": 11606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.657943455039397,
+      "learning_rate": 2.458713422081199e-05,
+      "loss": 0.9473,
+      "step": 11607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.563176507736924,
+      "learning_rate": 2.458296424434349e-05,
+      "loss": 1.6836,
+      "step": 11608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.465417336935041,
+      "learning_rate": 2.4578794279480923e-05,
+      "loss": 0.7192,
+      "step": 11609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.095548744184497,
+      "learning_rate": 2.4574624326340343e-05,
+      "loss": 1.1641,
+      "step": 11610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8673182485602848,
+      "learning_rate": 2.457045438503779e-05,
+      "loss": 1.3789,
+      "step": 11611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.045863839625963,
+      "learning_rate": 2.456628445568932e-05,
+      "loss": 1.2379,
+      "step": 11612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4317531751783488,
+      "learning_rate": 2.4562114538410975e-05,
+      "loss": 1.1777,
+      "step": 11613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3436547381655064,
+      "learning_rate": 2.4557944633318804e-05,
+      "loss": 1.1152,
+      "step": 11614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.318301993125991,
+      "learning_rate": 2.4553774740528848e-05,
+      "loss": 0.8125,
+      "step": 11615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.142899949928454,
+      "learning_rate": 2.4549604860157164e-05,
+      "loss": 1.4414,
+      "step": 11616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.214134573824005,
+      "learning_rate": 2.45454349923198e-05,
+      "loss": 1.1875,
+      "step": 11617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.018543155762913,
+      "learning_rate": 2.4541265137132787e-05,
+      "loss": 0.6416,
+      "step": 11618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4057901054484563,
+      "learning_rate": 2.4537095294712184e-05,
+      "loss": 0.7725,
+      "step": 11619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.720354067393368,
+      "learning_rate": 2.4532925465174027e-05,
+      "loss": 1.3711,
+      "step": 11620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4560760730212183,
+      "learning_rate": 2.452875564863437e-05,
+      "loss": 0.9397,
+      "step": 11621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4646814903963423,
+      "learning_rate": 2.4524585845209243e-05,
+      "loss": 1.0059,
+      "step": 11622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.534483919734964,
+      "learning_rate": 2.4520416055014706e-05,
+      "loss": 1.0059,
+      "step": 11623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1859143370889687,
+      "learning_rate": 2.4516246278166782e-05,
+      "loss": 0.9297,
+      "step": 11624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8446966424295737,
+      "learning_rate": 2.4512076514781536e-05,
+      "loss": 0.502,
+      "step": 11625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5750884698280574,
+      "learning_rate": 2.4507906764975e-05,
+      "loss": 1.082,
+      "step": 11626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.590377296116641,
+      "learning_rate": 2.4503737028863225e-05,
+      "loss": 0.7422,
+      "step": 11627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.321534794078403,
+      "learning_rate": 2.4499567306562238e-05,
+      "loss": 0.8333,
+      "step": 11628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.385454757808308,
+      "learning_rate": 2.4495397598188096e-05,
+      "loss": 1.2109,
+      "step": 11629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.179499712852179,
+      "learning_rate": 2.4491227903856828e-05,
+      "loss": 1.4453,
+      "step": 11630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.278314930783861,
+      "learning_rate": 2.4487058223684488e-05,
+      "loss": 1.0181,
+      "step": 11631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.780360685837742,
+      "learning_rate": 2.4482888557787103e-05,
+      "loss": 0.6094,
+      "step": 11632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9271119864114614,
+      "learning_rate": 2.447871890628072e-05,
+      "loss": 1.0225,
+      "step": 11633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.111600704753604,
+      "learning_rate": 2.447454926928137e-05,
+      "loss": 1.002,
+      "step": 11634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.797008873106078,
+      "learning_rate": 2.4470379646905108e-05,
+      "loss": 0.7559,
+      "step": 11635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.688718386525862,
+      "learning_rate": 2.4466210039267963e-05,
+      "loss": 1.4004,
+      "step": 11636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2551208079726073,
+      "learning_rate": 2.4462040446485974e-05,
+      "loss": 1.3652,
+      "step": 11637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0657684634798303,
+      "learning_rate": 2.445787086867519e-05,
+      "loss": 1.1855,
+      "step": 11638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.9227052775885936,
+      "learning_rate": 2.445370130595163e-05,
+      "loss": 0.9443,
+      "step": 11639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.27044774271126,
+      "learning_rate": 2.4449531758431344e-05,
+      "loss": 1.2832,
+      "step": 11640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.205764501745163,
+      "learning_rate": 2.444536222623036e-05,
+      "loss": 1.0083,
+      "step": 11641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.303967867479598,
+      "learning_rate": 2.4441192709464728e-05,
+      "loss": 1.4824,
+      "step": 11642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.08368548538766,
+      "learning_rate": 2.4437023208250463e-05,
+      "loss": 0.5337,
+      "step": 11643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5186117697522614,
+      "learning_rate": 2.4432853722703618e-05,
+      "loss": 0.3911,
+      "step": 11644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.783702212820498,
+      "learning_rate": 2.442868425294022e-05,
+      "loss": 0.7627,
+      "step": 11645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.757712386694278,
+      "learning_rate": 2.4424514799076315e-05,
+      "loss": 0.8772,
+      "step": 11646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.676277151516624,
+      "learning_rate": 2.4420345361227924e-05,
+      "loss": 0.6519,
+      "step": 11647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.6928247344906735,
+      "learning_rate": 2.441617593951109e-05,
+      "loss": 1.2227,
+      "step": 11648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.626381828271269,
+      "learning_rate": 2.4412006534041838e-05,
+      "loss": 0.8877,
+      "step": 11649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7131545799732792,
+      "learning_rate": 2.440783714493621e-05,
+      "loss": 1.0039,
+      "step": 11650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.170810575847966,
+      "learning_rate": 2.4403667772310223e-05,
+      "loss": 0.7449,
+      "step": 11651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4504000356631224,
+      "learning_rate": 2.4399498416279916e-05,
+      "loss": 0.6436,
+      "step": 11652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4482605929266774,
+      "learning_rate": 2.439532907696134e-05,
+      "loss": 1.041,
+      "step": 11653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.602172046998933,
+      "learning_rate": 2.43911597544705e-05,
+      "loss": 0.6943,
+      "step": 11654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.487568010330195,
+      "learning_rate": 2.4386990448923442e-05,
+      "loss": 1.4668,
+      "step": 11655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.787424995609382,
+      "learning_rate": 2.4382821160436182e-05,
+      "loss": 0.8515,
+      "step": 11656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3469472554988537,
+      "learning_rate": 2.437865188912477e-05,
+      "loss": 1.1157,
+      "step": 11657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9476469808603518,
+      "learning_rate": 2.437448263510521e-05,
+      "loss": 1.1338,
+      "step": 11658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.092942249868738,
+      "learning_rate": 2.4370313398493553e-05,
+      "loss": 1.2246,
+      "step": 11659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.407187796470855,
+      "learning_rate": 2.436614417940581e-05,
+      "loss": 0.7188,
+      "step": 11660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.083068595724458,
+      "learning_rate": 2.4361974977958014e-05,
+      "loss": 0.7764,
+      "step": 11661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.026602926853576,
+      "learning_rate": 2.4357805794266206e-05,
+      "loss": 1.0137,
+      "step": 11662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.833013595656916,
+      "learning_rate": 2.435363662844639e-05,
+      "loss": 1.8184,
+      "step": 11663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6052050135328355,
+      "learning_rate": 2.4349467480614617e-05,
+      "loss": 0.8721,
+      "step": 11664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.604912292606985,
+      "learning_rate": 2.4345298350886888e-05,
+      "loss": 0.8857,
+      "step": 11665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4220742420564245,
+      "learning_rate": 2.434112923937924e-05,
+      "loss": 1.4395,
+      "step": 11666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.267025895692084,
+      "learning_rate": 2.4336960146207704e-05,
+      "loss": 1.3594,
+      "step": 11667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.732365310623823,
+      "learning_rate": 2.4332791071488294e-05,
+      "loss": 1.062,
+      "step": 11668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9978415353642003,
+      "learning_rate": 2.432862201533704e-05,
+      "loss": 0.8284,
+      "step": 11669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4629419864087976,
+      "learning_rate": 2.4324452977869955e-05,
+      "loss": 1.3086,
+      "step": 11670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.756587573235043,
+      "learning_rate": 2.4320283959203062e-05,
+      "loss": 1.1309,
+      "step": 11671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.193347776314304,
+      "learning_rate": 2.4316114959452404e-05,
+      "loss": 1.0098,
+      "step": 11672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.608999049485601,
+      "learning_rate": 2.4311945978733984e-05,
+      "loss": 0.8584,
+      "step": 11673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0843929901948055,
+      "learning_rate": 2.430777701716383e-05,
+      "loss": 1.0889,
+      "step": 11674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.280253479970837,
+      "learning_rate": 2.4303608074857957e-05,
+      "loss": 0.8657,
+      "step": 11675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.747692133689716,
+      "learning_rate": 2.429943915193239e-05,
+      "loss": 1.3301,
+      "step": 11676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8313810878157883,
+      "learning_rate": 2.429527024850314e-05,
+      "loss": 1.3789,
+      "step": 11677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.061304943667612,
+      "learning_rate": 2.4291101364686242e-05,
+      "loss": 0.8906,
+      "step": 11678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5026542953061695,
+      "learning_rate": 2.4286932500597695e-05,
+      "loss": 1.4131,
+      "step": 11679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.47040068090805,
+      "learning_rate": 2.428276365635352e-05,
+      "loss": 1.0762,
+      "step": 11680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8266083249183214,
+      "learning_rate": 2.4278594832069756e-05,
+      "loss": 1.1152,
+      "step": 11681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6901762301211583,
+      "learning_rate": 2.4274426027862397e-05,
+      "loss": 1.2383,
+      "step": 11682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5549473061698724,
+      "learning_rate": 2.4270257243847473e-05,
+      "loss": 1.375,
+      "step": 11683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.476422316411107,
+      "learning_rate": 2.4266088480140984e-05,
+      "loss": 1.4707,
+      "step": 11684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.553529622805637,
+      "learning_rate": 2.4261919736858953e-05,
+      "loss": 1.0098,
+      "step": 11685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.719116361641585,
+      "learning_rate": 2.4257751014117406e-05,
+      "loss": 1.5137,
+      "step": 11686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.972521504419822,
+      "learning_rate": 2.4253582312032337e-05,
+      "loss": 0.6807,
+      "step": 11687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2163942475334717,
+      "learning_rate": 2.424941363071977e-05,
+      "loss": 0.7683,
+      "step": 11688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.930009152178593,
+      "learning_rate": 2.4245244970295713e-05,
+      "loss": 1.416,
+      "step": 11689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3389133161207343,
+      "learning_rate": 2.4241076330876184e-05,
+      "loss": 1.2793,
+      "step": 11690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.525819036219676,
+      "learning_rate": 2.4236907712577195e-05,
+      "loss": 1.2817,
+      "step": 11691
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0697239901449422,
+      "learning_rate": 2.4232739115514748e-05,
+      "loss": 0.7236,
+      "step": 11692
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.491553114173471,
+      "learning_rate": 2.4228570539804866e-05,
+      "loss": 1.064,
+      "step": 11693
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4363803774126347,
+      "learning_rate": 2.422440198556355e-05,
+      "loss": 1.5996,
+      "step": 11694
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4211603402545268,
+      "learning_rate": 2.4220233452906812e-05,
+      "loss": 0.9912,
+      "step": 11695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.989867609452047,
+      "learning_rate": 2.4216064941950655e-05,
+      "loss": 0.5972,
+      "step": 11696
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5954927771123697,
+      "learning_rate": 2.4211896452811098e-05,
+      "loss": 1.1543,
+      "step": 11697
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.054536420291411,
+      "learning_rate": 2.4207727985604128e-05,
+      "loss": 1.2207,
+      "step": 11698
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.11974599130902,
+      "learning_rate": 2.4203559540445775e-05,
+      "loss": 1.4707,
+      "step": 11699
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.734949977912457,
+      "learning_rate": 2.4199391117452038e-05,
+      "loss": 0.6709,
+      "step": 11700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7167763397296336,
+      "learning_rate": 2.4195222716738916e-05,
+      "loss": 1.0088,
+      "step": 11701
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8443486437563097,
+      "learning_rate": 2.4191054338422426e-05,
+      "loss": 1.1953,
+      "step": 11702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.096798526099636,
+      "learning_rate": 2.4186885982618557e-05,
+      "loss": 0.9995,
+      "step": 11703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.613224867139794,
+      "learning_rate": 2.4182717649443323e-05,
+      "loss": 0.8652,
+      "step": 11704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.102280908065858,
+      "learning_rate": 2.4178549339012723e-05,
+      "loss": 1.3105,
+      "step": 11705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.202138585239879,
+      "learning_rate": 2.417438105144276e-05,
+      "loss": 1.0557,
+      "step": 11706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4511503788586806,
+      "learning_rate": 2.4170212786849438e-05,
+      "loss": 1.2461,
+      "step": 11707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.650512889204481,
+      "learning_rate": 2.416604454534875e-05,
+      "loss": 1.5723,
+      "step": 11708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.415137236072566,
+      "learning_rate": 2.4161876327056705e-05,
+      "loss": 1.1748,
+      "step": 11709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2125672151532396,
+      "learning_rate": 2.4157708132089305e-05,
+      "loss": 0.8887,
+      "step": 11710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9200587180218878,
+      "learning_rate": 2.4153539960562544e-05,
+      "loss": 1.75,
+      "step": 11711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.364467503553329,
+      "learning_rate": 2.4149371812592422e-05,
+      "loss": 0.8896,
+      "step": 11712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8169482340580982,
+      "learning_rate": 2.4145203688294933e-05,
+      "loss": 1.5469,
+      "step": 11713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.94258262343261,
+      "learning_rate": 2.4141035587786083e-05,
+      "loss": 0.689,
+      "step": 11714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5997366698501363,
+      "learning_rate": 2.413686751118186e-05,
+      "loss": 0.7451,
+      "step": 11715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2171378927884544,
+      "learning_rate": 2.413269945859826e-05,
+      "loss": 1.0459,
+      "step": 11716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2342198458068028,
+      "learning_rate": 2.412853143015128e-05,
+      "loss": 0.981,
+      "step": 11717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0340030500075916,
+      "learning_rate": 2.4124363425956916e-05,
+      "loss": 0.833,
+      "step": 11718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4687242144433146,
+      "learning_rate": 2.4120195446131162e-05,
+      "loss": 1.3311,
+      "step": 11719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.579493119179055,
+      "learning_rate": 2.4116027490790014e-05,
+      "loss": 0.8682,
+      "step": 11720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.376488228334922,
+      "learning_rate": 2.4111859560049464e-05,
+      "loss": 1.3857,
+      "step": 11721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7344894072433115,
+      "learning_rate": 2.410769165402549e-05,
+      "loss": 1.6465,
+      "step": 11722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.76036857819071,
+      "learning_rate": 2.4103523772834105e-05,
+      "loss": 0.7637,
+      "step": 11723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0253371817472927,
+      "learning_rate": 2.409935591659128e-05,
+      "loss": 1.2051,
+      "step": 11724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7760806728409197,
+      "learning_rate": 2.4095188085413017e-05,
+      "loss": 0.8213,
+      "step": 11725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.358032552007479,
+      "learning_rate": 2.4091020279415293e-05,
+      "loss": 1.582,
+      "step": 11726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.654110205617211,
+      "learning_rate": 2.4086852498714117e-05,
+      "loss": 1.1133,
+      "step": 11727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.472733690286946,
+      "learning_rate": 2.4082684743425458e-05,
+      "loss": 1.4219,
+      "step": 11728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.447326954589197,
+      "learning_rate": 2.4078517013665314e-05,
+      "loss": 1.6621,
+      "step": 11729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8649310918969106,
+      "learning_rate": 2.4074349309549663e-05,
+      "loss": 0.8408,
+      "step": 11730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.563131419345681,
+      "learning_rate": 2.4070181631194495e-05,
+      "loss": 1.1934,
+      "step": 11731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9098834223517005,
+      "learning_rate": 2.406601397871579e-05,
+      "loss": 0.6214,
+      "step": 11732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.960338360804254,
+      "learning_rate": 2.4061846352229544e-05,
+      "loss": 1.373,
+      "step": 11733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2644808280741375,
+      "learning_rate": 2.4057678751851724e-05,
+      "loss": 0.7998,
+      "step": 11734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9058407731248597,
+      "learning_rate": 2.405351117769833e-05,
+      "loss": 0.5361,
+      "step": 11735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.419975963662319,
+      "learning_rate": 2.404934362988532e-05,
+      "loss": 1.2324,
+      "step": 11736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.66107170944758,
+      "learning_rate": 2.4045176108528694e-05,
+      "loss": 1.1855,
+      "step": 11737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8645170076957003,
+      "learning_rate": 2.4041008613744438e-05,
+      "loss": 1.3223,
+      "step": 11738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.568840648972932,
+      "learning_rate": 2.403684114564851e-05,
+      "loss": 0.6951,
+      "step": 11739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.642123720190772,
+      "learning_rate": 2.4032673704356912e-05,
+      "loss": 0.8965,
+      "step": 11740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.043531531633512,
+      "learning_rate": 2.4028506289985604e-05,
+      "loss": 1.3428,
+      "step": 11741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6982473449429314,
+      "learning_rate": 2.4024338902650574e-05,
+      "loss": 1.1514,
+      "step": 11742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.876231385941269,
+      "learning_rate": 2.4020171542467792e-05,
+      "loss": 0.6094,
+      "step": 11743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5042093033224626,
+      "learning_rate": 2.4016004209553243e-05,
+      "loss": 0.9692,
+      "step": 11744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2660748477635733,
+      "learning_rate": 2.401183690402288e-05,
+      "loss": 1.2705,
+      "step": 11745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.394408579137582,
+      "learning_rate": 2.4007669625992706e-05,
+      "loss": 0.8687,
+      "step": 11746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1607570383599546,
+      "learning_rate": 2.400350237557868e-05,
+      "loss": 0.4142,
+      "step": 11747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4845215634026245,
+      "learning_rate": 2.3999335152896784e-05,
+      "loss": 1.1025,
+      "step": 11748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3629657805644646,
+      "learning_rate": 2.3995167958062978e-05,
+      "loss": 1.2764,
+      "step": 11749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.52491206238438,
+      "learning_rate": 2.399100079119324e-05,
+      "loss": 1.2803,
+      "step": 11750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.616249568556279,
+      "learning_rate": 2.398683365240354e-05,
+      "loss": 0.9443,
+      "step": 11751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0452018643320264,
+      "learning_rate": 2.3982666541809846e-05,
+      "loss": 0.7637,
+      "step": 11752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.743102235255626,
+      "learning_rate": 2.3978499459528125e-05,
+      "loss": 1.0625,
+      "step": 11753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.191394894601891,
+      "learning_rate": 2.3974332405674344e-05,
+      "loss": 0.7661,
+      "step": 11754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.741718302641614,
+      "learning_rate": 2.3970165380364486e-05,
+      "loss": 1.209,
+      "step": 11755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8869405823827576,
+      "learning_rate": 2.3965998383714496e-05,
+      "loss": 0.8887,
+      "step": 11756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1238102667813545,
+      "learning_rate": 2.396183141584036e-05,
+      "loss": 0.6035,
+      "step": 11757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0786559910424613,
+      "learning_rate": 2.3957664476858025e-05,
+      "loss": 0.9355,
+      "step": 11758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6959042106952964,
+      "learning_rate": 2.395349756688347e-05,
+      "loss": 1.1543,
+      "step": 11759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.862216322106984,
+      "learning_rate": 2.3949330686032648e-05,
+      "loss": 1.2324,
+      "step": 11760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5535419473842174,
+      "learning_rate": 2.3945163834421526e-05,
+      "loss": 0.877,
+      "step": 11761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5321470013472527,
+      "learning_rate": 2.3940997012166065e-05,
+      "loss": 1.1387,
+      "step": 11762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.864365348860258,
+      "learning_rate": 2.3936830219382217e-05,
+      "loss": 1.2305,
+      "step": 11763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.007459480639381,
+      "learning_rate": 2.393266345618596e-05,
+      "loss": 1.0762,
+      "step": 11764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.504783903200008,
+      "learning_rate": 2.3928496722693246e-05,
+      "loss": 1.1641,
+      "step": 11765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9682857752793255,
+      "learning_rate": 2.3924330019020026e-05,
+      "loss": 0.9326,
+      "step": 11766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6139050084493385,
+      "learning_rate": 2.392016334528227e-05,
+      "loss": 0.9385,
+      "step": 11767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.800460652441113,
+      "learning_rate": 2.391599670159593e-05,
+      "loss": 1.1445,
+      "step": 11768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.299488285577978,
+      "learning_rate": 2.3911830088076958e-05,
+      "loss": 0.8594,
+      "step": 11769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1169238612592793,
+      "learning_rate": 2.390766350484131e-05,
+      "loss": 1.0527,
+      "step": 11770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.778279299040056,
+      "learning_rate": 2.3903496952004946e-05,
+      "loss": 1.4297,
+      "step": 11771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.049065699792437,
+      "learning_rate": 2.3899330429683807e-05,
+      "loss": 1.0625,
+      "step": 11772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6617002333704334,
+      "learning_rate": 2.3895163937993854e-05,
+      "loss": 0.9243,
+      "step": 11773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4813534578138965,
+      "learning_rate": 2.3890997477051047e-05,
+      "loss": 1.2173,
+      "step": 11774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.874215889856979,
+      "learning_rate": 2.3886831046971322e-05,
+      "loss": 1.1611,
+      "step": 11775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5783395388955332,
+      "learning_rate": 2.3882664647870645e-05,
+      "loss": 0.9946,
+      "step": 11776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9565926028683656,
+      "learning_rate": 2.3878498279864947e-05,
+      "loss": 1.0508,
+      "step": 11777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.442958375151143,
+      "learning_rate": 2.3874331943070194e-05,
+      "loss": 1.7559,
+      "step": 11778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.07262719467983,
+      "learning_rate": 2.3870165637602314e-05,
+      "loss": 0.8428,
+      "step": 11779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4210044034392353,
+      "learning_rate": 2.3865999363577277e-05,
+      "loss": 1.0762,
+      "step": 11780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.149913045322417,
+      "learning_rate": 2.3861833121111003e-05,
+      "loss": 1.3389,
+      "step": 11781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.761993836123855,
+      "learning_rate": 2.385766691031945e-05,
+      "loss": 0.7334,
+      "step": 11782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5367385323061225,
+      "learning_rate": 2.3853500731318568e-05,
+      "loss": 1.293,
+      "step": 11783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.250809311945027,
+      "learning_rate": 2.3849334584224294e-05,
+      "loss": 1.0527,
+      "step": 11784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.5829352928486236,
+      "learning_rate": 2.3845168469152568e-05,
+      "loss": 1.3521,
+      "step": 11785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.790126480052243,
+      "learning_rate": 2.3841002386219343e-05,
+      "loss": 1.0104,
+      "step": 11786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.4951256046396795,
+      "learning_rate": 2.3836836335540535e-05,
+      "loss": 1.123,
+      "step": 11787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5794446086026652,
+      "learning_rate": 2.383267031723211e-05,
+      "loss": 0.7568,
+      "step": 11788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.639753491669706,
+      "learning_rate": 2.3828504331409987e-05,
+      "loss": 0.4066,
+      "step": 11789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7389474113787897,
+      "learning_rate": 2.382433837819012e-05,
+      "loss": 0.709,
+      "step": 11790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.63528535418105,
+      "learning_rate": 2.382017245768842e-05,
+      "loss": 1.1963,
+      "step": 11791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.49435702465253,
+      "learning_rate": 2.381600657002085e-05,
+      "loss": 1.8574,
+      "step": 11792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2195218836595587,
+      "learning_rate": 2.3811840715303343e-05,
+      "loss": 0.8091,
+      "step": 11793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.006503072335586,
+      "learning_rate": 2.380767489365182e-05,
+      "loss": 1.1836,
+      "step": 11794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.203580086083068,
+      "learning_rate": 2.3803509105182222e-05,
+      "loss": 1.0625,
+      "step": 11795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1592899951682365,
+      "learning_rate": 2.379934335001047e-05,
+      "loss": 1.0391,
+      "step": 11796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9095931974183076,
+      "learning_rate": 2.3795177628252512e-05,
+      "loss": 0.9072,
+      "step": 11797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.500765841271792,
+      "learning_rate": 2.3791011940024266e-05,
+      "loss": 0.8877,
+      "step": 11798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6080952807361677,
+      "learning_rate": 2.378684628544167e-05,
+      "loss": 1.0449,
+      "step": 11799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.675777374750378,
+      "learning_rate": 2.378268066462064e-05,
+      "loss": 1.3711,
+      "step": 11800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.623589402394491,
+      "learning_rate": 2.377851507767711e-05,
+      "loss": 1.418,
+      "step": 11801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.144286761913078,
+      "learning_rate": 2.3774349524727012e-05,
+      "loss": 0.998,
+      "step": 11802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8093965861823265,
+      "learning_rate": 2.3770184005886277e-05,
+      "loss": 0.583,
+      "step": 11803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.495309411406911,
+      "learning_rate": 2.376601852127081e-05,
+      "loss": 1.166,
+      "step": 11804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.232635300040271,
+      "learning_rate": 2.3761853070996552e-05,
+      "loss": 0.7842,
+      "step": 11805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7857663004507316,
+      "learning_rate": 2.375768765517941e-05,
+      "loss": 0.8311,
+      "step": 11806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.661080579332563,
+      "learning_rate": 2.3753522273935326e-05,
+      "loss": 0.3848,
+      "step": 11807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.220088950161362,
+      "learning_rate": 2.3749356927380202e-05,
+      "loss": 1.416,
+      "step": 11808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9665975698360945,
+      "learning_rate": 2.3745191615629973e-05,
+      "loss": 1.6055,
+      "step": 11809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3900282370523303,
+      "learning_rate": 2.3741026338800537e-05,
+      "loss": 0.873,
+      "step": 11810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3914830282423787,
+      "learning_rate": 2.373686109700783e-05,
+      "loss": 1.1499,
+      "step": 11811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.139768592968364,
+      "learning_rate": 2.3732695890367773e-05,
+      "loss": 1.5078,
+      "step": 11812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2708515784787635,
+      "learning_rate": 2.3728530718996267e-05,
+      "loss": 0.7554,
+      "step": 11813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.357186815733504,
+      "learning_rate": 2.3724365583009237e-05,
+      "loss": 1.2891,
+      "step": 11814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.84554711871833,
+      "learning_rate": 2.3720200482522588e-05,
+      "loss": 1.3027,
+      "step": 11815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.827577906849857,
+      "learning_rate": 2.3716035417652248e-05,
+      "loss": 0.9434,
+      "step": 11816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.143324771446739,
+      "learning_rate": 2.3711870388514108e-05,
+      "loss": 1.5098,
+      "step": 11817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.575717796351703,
+      "learning_rate": 2.3707705395224103e-05,
+      "loss": 1.2734,
+      "step": 11818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.486113633245987,
+      "learning_rate": 2.3703540437898118e-05,
+      "loss": 1.4014,
+      "step": 11819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.62085405552793,
+      "learning_rate": 2.3699375516652077e-05,
+      "loss": 1.0376,
+      "step": 11820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.611564467573905,
+      "learning_rate": 2.3695210631601895e-05,
+      "loss": 1.1475,
+      "step": 11821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3686802755677623,
+      "learning_rate": 2.369104578286346e-05,
+      "loss": 1.2588,
+      "step": 11822
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.155251534423707,
+      "learning_rate": 2.3686880970552693e-05,
+      "loss": 1.1777,
+      "step": 11823
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.161495245818543,
+      "learning_rate": 2.3682716194785496e-05,
+      "loss": 1.3418,
+      "step": 11824
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.307913732601787,
+      "learning_rate": 2.3678551455677766e-05,
+      "loss": 1.2617,
+      "step": 11825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.307527516683815,
+      "learning_rate": 2.3674386753345418e-05,
+      "loss": 1.498,
+      "step": 11826
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.566243067131806,
+      "learning_rate": 2.3670222087904343e-05,
+      "loss": 0.9207,
+      "step": 11827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.444943043644129,
+      "learning_rate": 2.3666057459470436e-05,
+      "loss": 1.0312,
+      "step": 11828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2849138701919505,
+      "learning_rate": 2.3661892868159622e-05,
+      "loss": 1.2031,
+      "step": 11829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.27221148825362,
+      "learning_rate": 2.3657728314087778e-05,
+      "loss": 1.1367,
+      "step": 11830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3560323008823127,
+      "learning_rate": 2.365356379737081e-05,
+      "loss": 1.5762,
+      "step": 11831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9241804056878267,
+      "learning_rate": 2.3649399318124614e-05,
+      "loss": 1.8184,
+      "step": 11832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4838617742296374,
+      "learning_rate": 2.3645234876465087e-05,
+      "loss": 0.748,
+      "step": 11833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.679993087334523,
+      "learning_rate": 2.3641070472508118e-05,
+      "loss": 1.3965,
+      "step": 11834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9216000024647206,
+      "learning_rate": 2.363690610636961e-05,
+      "loss": 0.5264,
+      "step": 11835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8381925502350036,
+      "learning_rate": 2.3632741778165442e-05,
+      "loss": 1.2002,
+      "step": 11836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.517253182973203,
+      "learning_rate": 2.362857748801152e-05,
+      "loss": 1.3535,
+      "step": 11837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0067255662699703,
+      "learning_rate": 2.362441323602372e-05,
+      "loss": 0.9854,
+      "step": 11838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.090613677902241,
+      "learning_rate": 2.3620249022317945e-05,
+      "loss": 1.0161,
+      "step": 11839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3274733214308396,
+      "learning_rate": 2.361608484701008e-05,
+      "loss": 1.1001,
+      "step": 11840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.397330829599477,
+      "learning_rate": 2.3611920710216006e-05,
+      "loss": 0.7998,
+      "step": 11841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3269576770141747,
+      "learning_rate": 2.3607756612051616e-05,
+      "loss": 0.9189,
+      "step": 11842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.074377655464003,
+      "learning_rate": 2.3603592552632796e-05,
+      "loss": 1.5996,
+      "step": 11843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.232885348022806,
+      "learning_rate": 2.3599428532075424e-05,
+      "loss": 1.1406,
+      "step": 11844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.842789444653764,
+      "learning_rate": 2.359526455049539e-05,
+      "loss": 0.8506,
+      "step": 11845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.794217156646033,
+      "learning_rate": 2.3591100608008562e-05,
+      "loss": 0.6328,
+      "step": 11846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8063421017628025,
+      "learning_rate": 2.358693670473083e-05,
+      "loss": 1.5234,
+      "step": 11847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9204644863201015,
+      "learning_rate": 2.3582772840778084e-05,
+      "loss": 0.9023,
+      "step": 11848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.297606572450473,
+      "learning_rate": 2.3578609016266186e-05,
+      "loss": 1.9746,
+      "step": 11849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.099306915083086,
+      "learning_rate": 2.3574445231311026e-05,
+      "loss": 1.0723,
+      "step": 11850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.4213622869780504,
+      "learning_rate": 2.3570281486028472e-05,
+      "loss": 1.327,
+      "step": 11851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.773659041090753,
+      "learning_rate": 2.3566117780534404e-05,
+      "loss": 0.8438,
+      "step": 11852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9842626219216393,
+      "learning_rate": 2.3561954114944692e-05,
+      "loss": 1.1465,
+      "step": 11853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1849272721536592,
+      "learning_rate": 2.3557790489375216e-05,
+      "loss": 1.3418,
+      "step": 11854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.25372981046223,
+      "learning_rate": 2.355362690394184e-05,
+      "loss": 0.7935,
+      "step": 11855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9024881606801056,
+      "learning_rate": 2.354946335876043e-05,
+      "loss": 0.6904,
+      "step": 11856
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.879586612466639,
+      "learning_rate": 2.3545299853946874e-05,
+      "loss": 0.9824,
+      "step": 11857
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9236066506797347,
+      "learning_rate": 2.3541136389617032e-05,
+      "loss": 0.9287,
+      "step": 11858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.020281095195793,
+      "learning_rate": 2.353697296588677e-05,
+      "loss": 0.8955,
+      "step": 11859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4295541413926656,
+      "learning_rate": 2.353280958287195e-05,
+      "loss": 1.2754,
+      "step": 11860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.875057904804343,
+      "learning_rate": 2.3528646240688447e-05,
+      "loss": 0.9358,
+      "step": 11861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4850653404995535,
+      "learning_rate": 2.3524482939452118e-05,
+      "loss": 0.834,
+      "step": 11862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.717671628258738,
+      "learning_rate": 2.3520319679278825e-05,
+      "loss": 1.7324,
+      "step": 11863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.265667782735976,
+      "learning_rate": 2.351615646028444e-05,
+      "loss": 0.6936,
+      "step": 11864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.333108724984197,
+      "learning_rate": 2.3511993282584803e-05,
+      "loss": 1.1465,
+      "step": 11865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7476327417847872,
+      "learning_rate": 2.3507830146295794e-05,
+      "loss": 1.0815,
+      "step": 11866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2122771716652596,
+      "learning_rate": 2.3503667051533274e-05,
+      "loss": 0.5254,
+      "step": 11867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9576713444148177,
+      "learning_rate": 2.349950399841308e-05,
+      "loss": 0.4836,
+      "step": 11868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.558600444510937,
+      "learning_rate": 2.3495340987051086e-05,
+      "loss": 1.0527,
+      "step": 11869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.378828237905597,
+      "learning_rate": 2.3491178017563136e-05,
+      "loss": 1.3154,
+      "step": 11870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8993259438910677,
+      "learning_rate": 2.3487015090065093e-05,
+      "loss": 0.7685,
+      "step": 11871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.017598193450662,
+      "learning_rate": 2.3482852204672797e-05,
+      "loss": 1.0332,
+      "step": 11872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.22882709973799,
+      "learning_rate": 2.347868936150211e-05,
+      "loss": 1.6016,
+      "step": 11873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1178164504898795,
+      "learning_rate": 2.3474526560668877e-05,
+      "loss": 1.3496,
+      "step": 11874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.338815204962352,
+      "learning_rate": 2.347036380228894e-05,
+      "loss": 0.9014,
+      "step": 11875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.373768436948861,
+      "learning_rate": 2.346620108647817e-05,
+      "loss": 1.3535,
+      "step": 11876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.285244514521346,
+      "learning_rate": 2.3462038413352394e-05,
+      "loss": 1.4463,
+      "step": 11877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.729145861887345,
+      "learning_rate": 2.3457875783027468e-05,
+      "loss": 1.0371,
+      "step": 11878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.640140285810541,
+      "learning_rate": 2.3453713195619224e-05,
+      "loss": 0.9453,
+      "step": 11879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2900170325212277,
+      "learning_rate": 2.3449550651243517e-05,
+      "loss": 0.9058,
+      "step": 11880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.221362228349112,
+      "learning_rate": 2.3445388150016177e-05,
+      "loss": 0.1873,
+      "step": 11881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.833057801556122,
+      "learning_rate": 2.344122569205306e-05,
+      "loss": 0.7295,
+      "step": 11882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2338247622043466,
+      "learning_rate": 2.3437063277469994e-05,
+      "loss": 1.0039,
+      "step": 11883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8307998400407968,
+      "learning_rate": 2.3432900906382815e-05,
+      "loss": 1.5449,
+      "step": 11884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1650987478382193,
+      "learning_rate": 2.342873857890737e-05,
+      "loss": 0.9053,
+      "step": 11885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.305758971191161,
+      "learning_rate": 2.3424576295159497e-05,
+      "loss": 1.166,
+      "step": 11886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.961778989202906,
+      "learning_rate": 2.3420414055255017e-05,
+      "loss": 1.3682,
+      "step": 11887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.475532293374648,
+      "learning_rate": 2.3416251859309776e-05,
+      "loss": 0.9707,
+      "step": 11888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6433075590989685,
+      "learning_rate": 2.3412089707439596e-05,
+      "loss": 0.9531,
+      "step": 11889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.496987924224947,
+      "learning_rate": 2.340792759976032e-05,
+      "loss": 1.0176,
+      "step": 11890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7105427655250587,
+      "learning_rate": 2.3403765536387763e-05,
+      "loss": 1.082,
+      "step": 11891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.412997916055065,
+      "learning_rate": 2.339960351743777e-05,
+      "loss": 0.6338,
+      "step": 11892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3192869509937086,
+      "learning_rate": 2.3395441543026146e-05,
+      "loss": 1.0029,
+      "step": 11893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.031829015580047,
+      "learning_rate": 2.3391279613268737e-05,
+      "loss": 1.3184,
+      "step": 11894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7862759729529367,
+      "learning_rate": 2.338711772828137e-05,
+      "loss": 0.9458,
+      "step": 11895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.798203098686223,
+      "learning_rate": 2.338295588817985e-05,
+      "loss": 1.0427,
+      "step": 11896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4430231286124346,
+      "learning_rate": 2.3378794093080014e-05,
+      "loss": 1.3809,
+      "step": 11897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.064215553494315,
+      "learning_rate": 2.3374632343097677e-05,
+      "loss": 0.8125,
+      "step": 11898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7295595668910133,
+      "learning_rate": 2.337047063834866e-05,
+      "loss": 1.0713,
+      "step": 11899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.36295928723446,
+      "learning_rate": 2.3366308978948783e-05,
+      "loss": 0.9043,
+      "step": 11900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.026391947766633,
+      "learning_rate": 2.3362147365013866e-05,
+      "loss": 1.5059,
+      "step": 11901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8530508677283017,
+      "learning_rate": 2.3357985796659703e-05,
+      "loss": 1.1699,
+      "step": 11902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.817286897389216,
+      "learning_rate": 2.3353824274002144e-05,
+      "loss": 0.7505,
+      "step": 11903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.275923794997536,
+      "learning_rate": 2.3349662797156978e-05,
+      "loss": 1.1514,
+      "step": 11904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0069556663094303,
+      "learning_rate": 2.334550136624003e-05,
+      "loss": 0.8398,
+      "step": 11905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9757684840490053,
+      "learning_rate": 2.3341339981367104e-05,
+      "loss": 0.8945,
+      "step": 11906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.06567902901389,
+      "learning_rate": 2.333717864265401e-05,
+      "loss": 1.2129,
+      "step": 11907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3759571957377057,
+      "learning_rate": 2.3333017350216558e-05,
+      "loss": 1.543,
+      "step": 11908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.261654551489289,
+      "learning_rate": 2.3328856104170557e-05,
+      "loss": 1.248,
+      "step": 11909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3455630601266644,
+      "learning_rate": 2.3324694904631805e-05,
+      "loss": 1.1133,
+      "step": 11910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4534786267688578,
+      "learning_rate": 2.332053375171612e-05,
+      "loss": 1.0,
+      "step": 11911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7393339693903176,
+      "learning_rate": 2.3316372645539282e-05,
+      "loss": 1.0273,
+      "step": 11912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1684059595743546,
+      "learning_rate": 2.3312211586217117e-05,
+      "loss": 0.625,
+      "step": 11913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2522382200078894,
+      "learning_rate": 2.330805057386542e-05,
+      "loss": 0.7114,
+      "step": 11914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6766399704143033,
+      "learning_rate": 2.330388960859998e-05,
+      "loss": 0.8564,
+      "step": 11915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.321965538829491,
+      "learning_rate": 2.329972869053661e-05,
+      "loss": 1.5781,
+      "step": 11916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.351954836132806,
+      "learning_rate": 2.329556781979109e-05,
+      "loss": 0.8467,
+      "step": 11917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8958399564452324,
+      "learning_rate": 2.329140699647923e-05,
+      "loss": 1.4082,
+      "step": 11918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.034180039118729,
+      "learning_rate": 2.3287246220716815e-05,
+      "loss": 1.0615,
+      "step": 11919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4193697862265915,
+      "learning_rate": 2.3283085492619644e-05,
+      "loss": 0.6963,
+      "step": 11920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.451781761619518,
+      "learning_rate": 2.327892481230349e-05,
+      "loss": 0.7097,
+      "step": 11921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.60504298551463,
+      "learning_rate": 2.327476417988417e-05,
+      "loss": 1.2598,
+      "step": 11922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0258454241667074,
+      "learning_rate": 2.327060359547746e-05,
+      "loss": 0.7451,
+      "step": 11923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.195063578290956,
+      "learning_rate": 2.326644305919915e-05,
+      "loss": 0.5657,
+      "step": 11924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375635664348155,
+      "learning_rate": 2.3262282571165016e-05,
+      "loss": 1.0117,
+      "step": 11925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5342382993884294,
+      "learning_rate": 2.325812213149086e-05,
+      "loss": 0.9238,
+      "step": 11926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.002204807812846,
+      "learning_rate": 2.3253961740292452e-05,
+      "loss": 1.4922,
+      "step": 11927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.057287536546493,
+      "learning_rate": 2.3249801397685582e-05,
+      "loss": 1.0303,
+      "step": 11928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.59428413036593,
+      "learning_rate": 2.3245641103786017e-05,
+      "loss": 1.457,
+      "step": 11929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1075027426525725,
+      "learning_rate": 2.3241480858709543e-05,
+      "loss": 0.8564,
+      "step": 11930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.284871610210623,
+      "learning_rate": 2.3237320662571953e-05,
+      "loss": 0.8086,
+      "step": 11931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.193055640658462,
+      "learning_rate": 2.323316051548901e-05,
+      "loss": 1.0039,
+      "step": 11932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9458795050535005,
+      "learning_rate": 2.3229000417576486e-05,
+      "loss": 1.3398,
+      "step": 11933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.114385321127923,
+      "learning_rate": 2.322484036895016e-05,
+      "loss": 0.8112,
+      "step": 11934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.3269397838090065,
+      "learning_rate": 2.322068036972581e-05,
+      "loss": 1.2783,
+      "step": 11935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.803607369729808,
+      "learning_rate": 2.3216520420019195e-05,
+      "loss": 0.9677,
+      "step": 11936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.588121081612445,
+      "learning_rate": 2.3212360519946098e-05,
+      "loss": 0.9893,
+      "step": 11937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.00342504958915,
+      "learning_rate": 2.320820066962227e-05,
+      "loss": 1.1899,
+      "step": 11938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7668041108522874,
+      "learning_rate": 2.3204040869163485e-05,
+      "loss": 0.8701,
+      "step": 11939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.144049990540117,
+      "learning_rate": 2.3199881118685514e-05,
+      "loss": 1.2559,
+      "step": 11940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.9015122396162,
+      "learning_rate": 2.3195721418304127e-05,
+      "loss": 1.3438,
+      "step": 11941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.328530443595256,
+      "learning_rate": 2.3191561768135074e-05,
+      "loss": 1.2119,
+      "step": 11942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.816428822559637,
+      "learning_rate": 2.3187402168294123e-05,
+      "loss": 1.3223,
+      "step": 11943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6963991490169055,
+      "learning_rate": 2.3183242618897027e-05,
+      "loss": 1.6484,
+      "step": 11944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.717377151618259,
+      "learning_rate": 2.3179083120059558e-05,
+      "loss": 1.0254,
+      "step": 11945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.532077056121945,
+      "learning_rate": 2.3174923671897454e-05,
+      "loss": 1.0977,
+      "step": 11946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7002891703567866,
+      "learning_rate": 2.3170764274526492e-05,
+      "loss": 1.1816,
+      "step": 11947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.125017491752734,
+      "learning_rate": 2.3166604928062406e-05,
+      "loss": 1.9531,
+      "step": 11948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4781519310415,
+      "learning_rate": 2.3162445632620956e-05,
+      "loss": 1.0801,
+      "step": 11949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5818495232002228,
+      "learning_rate": 2.3158286388317904e-05,
+      "loss": 1.0498,
+      "step": 11950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.263362969441389,
+      "learning_rate": 2.315412719526899e-05,
+      "loss": 1.5283,
+      "step": 11951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.121911482110869,
+      "learning_rate": 2.3149968053589965e-05,
+      "loss": 0.9082,
+      "step": 11952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2455389515215036,
+      "learning_rate": 2.3145808963396575e-05,
+      "loss": 0.9883,
+      "step": 11953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.410135980935,
+      "learning_rate": 2.314164992480457e-05,
+      "loss": 1.2012,
+      "step": 11954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.95703802271764,
+      "learning_rate": 2.313749093792969e-05,
+      "loss": 1.2939,
+      "step": 11955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.977324946667581,
+      "learning_rate": 2.313333200288768e-05,
+      "loss": 1.7383,
+      "step": 11956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.073720483198299,
+      "learning_rate": 2.3129173119794272e-05,
+      "loss": 1.6211,
+      "step": 11957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7248194554758105,
+      "learning_rate": 2.3125014288765215e-05,
+      "loss": 1.4297,
+      "step": 11958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.845545107840393,
+      "learning_rate": 2.312085550991625e-05,
+      "loss": 1.6621,
+      "step": 11959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8786653931349224,
+      "learning_rate": 2.3116696783363118e-05,
+      "loss": 1.0322,
+      "step": 11960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5008443360265127,
+      "learning_rate": 2.311253810922154e-05,
+      "loss": 0.7285,
+      "step": 11961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.543799011358163,
+      "learning_rate": 2.3108379487607265e-05,
+      "loss": 0.9941,
+      "step": 11962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.0506602167318615,
+      "learning_rate": 2.3104220918636013e-05,
+      "loss": 0.877,
+      "step": 11963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0948539214998014,
+      "learning_rate": 2.3100062402423526e-05,
+      "loss": 0.7942,
+      "step": 11964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.280260704424494,
+      "learning_rate": 2.3095903939085522e-05,
+      "loss": 1.1602,
+      "step": 11965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1467073999809134,
+      "learning_rate": 2.3091745528737747e-05,
+      "loss": 1.2217,
+      "step": 11966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7688457463847547,
+      "learning_rate": 2.30875871714959e-05,
+      "loss": 1.1963,
+      "step": 11967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.267135116844827,
+      "learning_rate": 2.308342886747573e-05,
+      "loss": 1.5391,
+      "step": 11968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.388200165295348,
+      "learning_rate": 2.3079270616792966e-05,
+      "loss": 1.1416,
+      "step": 11969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4373790026246027,
+      "learning_rate": 2.3075112419563308e-05,
+      "loss": 0.8164,
+      "step": 11970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8632214632336725,
+      "learning_rate": 2.3070954275902494e-05,
+      "loss": 0.7744,
+      "step": 11971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.188262414560799,
+      "learning_rate": 2.3066796185926232e-05,
+      "loss": 1.1875,
+      "step": 11972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1328660765368865,
+      "learning_rate": 2.3062638149750247e-05,
+      "loss": 1.7012,
+      "step": 11973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0304360739708547,
+      "learning_rate": 2.3058480167490253e-05,
+      "loss": 1.1436,
+      "step": 11974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.75154815041809,
+      "learning_rate": 2.3054322239261973e-05,
+      "loss": 0.666,
+      "step": 11975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8695744093658595,
+      "learning_rate": 2.30501643651811e-05,
+      "loss": 0.748,
+      "step": 11976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.339781796967802,
+      "learning_rate": 2.304600654536336e-05,
+      "loss": 0.54,
+      "step": 11977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7239269319038957,
+      "learning_rate": 2.304184877992447e-05,
+      "loss": 1.0054,
+      "step": 11978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.3459368241070475,
+      "learning_rate": 2.3037691068980127e-05,
+      "loss": 0.6582,
+      "step": 11979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1882112709086368,
+      "learning_rate": 2.3033533412646045e-05,
+      "loss": 1.0195,
+      "step": 11980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.857945468048425,
+      "learning_rate": 2.3029375811037928e-05,
+      "loss": 1.0801,
+      "step": 11981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.908147275045944,
+      "learning_rate": 2.3025218264271478e-05,
+      "loss": 0.7761,
+      "step": 11982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3709977207502706,
+      "learning_rate": 2.3021060772462406e-05,
+      "loss": 0.8813,
+      "step": 11983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.623803539494215,
+      "learning_rate": 2.30169033357264e-05,
+      "loss": 2.1211,
+      "step": 11984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.793396816928263,
+      "learning_rate": 2.301274595417917e-05,
+      "loss": 1.0391,
+      "step": 11985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.734499485909791,
+      "learning_rate": 2.30085886279364e-05,
+      "loss": 1.3975,
+      "step": 11986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.448625154863589,
+      "learning_rate": 2.3004431357113808e-05,
+      "loss": 0.9082,
+      "step": 11987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3971077777374967,
+      "learning_rate": 2.300027414182708e-05,
+      "loss": 1.3706,
+      "step": 11988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.360284289031478,
+      "learning_rate": 2.2996116982191906e-05,
+      "loss": 1.3906,
+      "step": 11989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7668680379147177,
+      "learning_rate": 2.2991959878323985e-05,
+      "loss": 1.0342,
+      "step": 11990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.712484643564028,
+      "learning_rate": 2.2987802830338998e-05,
+      "loss": 0.8896,
+      "step": 11991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3495438905770953,
+      "learning_rate": 2.2983645838352643e-05,
+      "loss": 0.7812,
+      "step": 11992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.813134100746091,
+      "learning_rate": 2.29794889024806e-05,
+      "loss": 1.0762,
+      "step": 11993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.045087660050904,
+      "learning_rate": 2.2975332022838564e-05,
+      "loss": 1.3652,
+      "step": 11994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.712161085283841,
+      "learning_rate": 2.29711751995422e-05,
+      "loss": 1.2334,
+      "step": 11995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.276310162304052,
+      "learning_rate": 2.2967018432707213e-05,
+      "loss": 1.1377,
+      "step": 11996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6077708425213364,
+      "learning_rate": 2.2962861722449278e-05,
+      "loss": 0.3249,
+      "step": 11997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.053008038576608,
+      "learning_rate": 2.2958705068884068e-05,
+      "loss": 1.6348,
+      "step": 11998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.333944146623749,
+      "learning_rate": 2.2954548472127264e-05,
+      "loss": 1.3579,
+      "step": 11999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.833010000591623,
+      "learning_rate": 2.295039193229455e-05,
+      "loss": 0.7256,
+      "step": 12000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.013440271447,
+      "learning_rate": 2.2946235449501587e-05,
+      "loss": 0.8574,
+      "step": 12001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.483645921514289,
+      "learning_rate": 2.2942079023864063e-05,
+      "loss": 0.7646,
+      "step": 12002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7332583238241557,
+      "learning_rate": 2.2937922655497635e-05,
+      "loss": 1.0518,
+      "step": 12003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.316257111747621,
+      "learning_rate": 2.2933766344517977e-05,
+      "loss": 1.4434,
+      "step": 12004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0992781124942477,
+      "learning_rate": 2.2929610091040767e-05,
+      "loss": 1.0195,
+      "step": 12005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.460708369469763,
+      "learning_rate": 2.2925453895181662e-05,
+      "loss": 0.8828,
+      "step": 12006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7475856065860698,
+      "learning_rate": 2.2921297757056336e-05,
+      "loss": 0.8159,
+      "step": 12007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2738823209725143,
+      "learning_rate": 2.2917141676780443e-05,
+      "loss": 0.8398,
+      "step": 12008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0624579601905673,
+      "learning_rate": 2.2912985654469654e-05,
+      "loss": 0.6987,
+      "step": 12009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.292685724149668,
+      "learning_rate": 2.290882969023962e-05,
+      "loss": 1.0474,
+      "step": 12010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.705698263646725,
+      "learning_rate": 2.2904673784206007e-05,
+      "loss": 0.9307,
+      "step": 12011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3347789172966316,
+      "learning_rate": 2.2900517936484467e-05,
+      "loss": 1.418,
+      "step": 12012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7571432108467095,
+      "learning_rate": 2.2896362147190663e-05,
+      "loss": 1.2051,
+      "step": 12013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.180117323403012,
+      "learning_rate": 2.2892206416440232e-05,
+      "loss": 1.6172,
+      "step": 12014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3870384888696226,
+      "learning_rate": 2.2888050744348842e-05,
+      "loss": 1.1851,
+      "step": 12015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6154291810443437,
+      "learning_rate": 2.2883895131032147e-05,
+      "loss": 0.8633,
+      "step": 12016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6618579340763495,
+      "learning_rate": 2.2879739576605787e-05,
+      "loss": 1.0723,
+      "step": 12017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.398322087098723,
+      "learning_rate": 2.2875584081185412e-05,
+      "loss": 0.313,
+      "step": 12018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2691755203826007,
+      "learning_rate": 2.2871428644886663e-05,
+      "loss": 0.9907,
+      "step": 12019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6960298772169486,
+      "learning_rate": 2.286727326782519e-05,
+      "loss": 1.5469,
+      "step": 12020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.819553468466784,
+      "learning_rate": 2.2863117950116638e-05,
+      "loss": 1.248,
+      "step": 12021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.7432699209349956,
+      "learning_rate": 2.285896269187664e-05,
+      "loss": 0.6133,
+      "step": 12022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.591870396981581,
+      "learning_rate": 2.285480749322083e-05,
+      "loss": 0.7305,
+      "step": 12023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1894167540606393,
+      "learning_rate": 2.285065235426487e-05,
+      "loss": 1.4053,
+      "step": 12024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.903741859044349,
+      "learning_rate": 2.284649727512437e-05,
+      "loss": 1.7383,
+      "step": 12025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.102606379109178,
+      "learning_rate": 2.284234225591498e-05,
+      "loss": 0.7949,
+      "step": 12026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7145465496046857,
+      "learning_rate": 2.2838187296752323e-05,
+      "loss": 0.9541,
+      "step": 12027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2849856581957795,
+      "learning_rate": 2.2834032397752038e-05,
+      "loss": 0.6724,
+      "step": 12028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0707751638837832,
+      "learning_rate": 2.2829877559029745e-05,
+      "loss": 0.5714,
+      "step": 12029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.800067646821501,
+      "learning_rate": 2.282572278070108e-05,
+      "loss": 0.6074,
+      "step": 12030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0508518967892524,
+      "learning_rate": 2.282156806288166e-05,
+      "loss": 1.4336,
+      "step": 12031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.928258766204836,
+      "learning_rate": 2.2817413405687106e-05,
+      "loss": 0.5774,
+      "step": 12032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8848293293466387,
+      "learning_rate": 2.281325880923306e-05,
+      "loss": 1.1023,
+      "step": 12033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2711364007231483,
+      "learning_rate": 2.2809104273635128e-05,
+      "loss": 0.8389,
+      "step": 12034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.853692075360243,
+      "learning_rate": 2.280494979900894e-05,
+      "loss": 1.1113,
+      "step": 12035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6763232942083195,
+      "learning_rate": 2.2800795385470098e-05,
+      "loss": 1.1191,
+      "step": 12036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.565007298691592,
+      "learning_rate": 2.2796641033134228e-05,
+      "loss": 1.541,
+      "step": 12037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2165341124901,
+      "learning_rate": 2.2792486742116937e-05,
+      "loss": 0.6877,
+      "step": 12038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.578338692064263,
+      "learning_rate": 2.278833251253384e-05,
+      "loss": 1.4727,
+      "step": 12039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8225881425256962,
+      "learning_rate": 2.2784178344500556e-05,
+      "loss": 1.6973,
+      "step": 12040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6932587995709034,
+      "learning_rate": 2.278002423813268e-05,
+      "loss": 1.2832,
+      "step": 12041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1199059857809273,
+      "learning_rate": 2.2775870193545823e-05,
+      "loss": 1.1611,
+      "step": 12042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4339731330153827,
+      "learning_rate": 2.2771716210855603e-05,
+      "loss": 1.0891,
+      "step": 12043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.346510215478555,
+      "learning_rate": 2.276756229017761e-05,
+      "loss": 0.5248,
+      "step": 12044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.947666178526784,
+      "learning_rate": 2.276340843162745e-05,
+      "loss": 1.3125,
+      "step": 12045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.550881633565045,
+      "learning_rate": 2.275925463532072e-05,
+      "loss": 1.3154,
+      "step": 12046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.22872017675694,
+      "learning_rate": 2.2755100901373026e-05,
+      "loss": 1.4502,
+      "step": 12047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.163344248166419,
+      "learning_rate": 2.2750947229899956e-05,
+      "loss": 0.9062,
+      "step": 12048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.561943546136954,
+      "learning_rate": 2.2746793621017114e-05,
+      "loss": 0.5459,
+      "step": 12049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1671721071792325,
+      "learning_rate": 2.274264007484008e-05,
+      "loss": 1.1016,
+      "step": 12050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.42281226821921,
+      "learning_rate": 2.2738486591484453e-05,
+      "loss": 0.7446,
+      "step": 12051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.467477788544621,
+      "learning_rate": 2.2734333171065833e-05,
+      "loss": 1.1953,
+      "step": 12052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.178051720459248,
+      "learning_rate": 2.2730179813699792e-05,
+      "loss": 0.8984,
+      "step": 12053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4811088156709684,
+      "learning_rate": 2.272602651950193e-05,
+      "loss": 1.541,
+      "step": 12054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1164662163604495,
+      "learning_rate": 2.2721873288587814e-05,
+      "loss": 1.3184,
+      "step": 12055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4308488529596115,
+      "learning_rate": 2.2717720121073045e-05,
+      "loss": 0.7607,
+      "step": 12056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2580383610199464,
+      "learning_rate": 2.271356701707319e-05,
+      "loss": 0.9219,
+      "step": 12057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.14349795304224,
+      "learning_rate": 2.2709413976703844e-05,
+      "loss": 1.168,
+      "step": 12058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6118303271122514,
+      "learning_rate": 2.2705261000080567e-05,
+      "loss": 0.8926,
+      "step": 12059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7283749441440426,
+      "learning_rate": 2.2701108087318933e-05,
+      "loss": 1.1807,
+      "step": 12060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.408884390695527,
+      "learning_rate": 2.2696955238534534e-05,
+      "loss": 0.9326,
+      "step": 12061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5691761415174668,
+      "learning_rate": 2.2692802453842936e-05,
+      "loss": 0.5121,
+      "step": 12062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.964301271012367,
+      "learning_rate": 2.2688649733359705e-05,
+      "loss": 0.9814,
+      "step": 12063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.653782596161027,
+      "learning_rate": 2.268449707720041e-05,
+      "loss": 1.3906,
+      "step": 12064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.258454756152888,
+      "learning_rate": 2.268034448548062e-05,
+      "loss": 0.7412,
+      "step": 12065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8989656443689686,
+      "learning_rate": 2.2676191958315902e-05,
+      "loss": 0.9126,
+      "step": 12066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.235366044240603,
+      "learning_rate": 2.267203949582181e-05,
+      "loss": 1.186,
+      "step": 12067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.66777928967034,
+      "learning_rate": 2.2667887098113915e-05,
+      "loss": 0.7695,
+      "step": 12068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.773518439152708,
+      "learning_rate": 2.266373476530776e-05,
+      "loss": 1.2178,
+      "step": 12069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1663596840554162,
+      "learning_rate": 2.2659582497518926e-05,
+      "loss": 0.3958,
+      "step": 12070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.025762289915204,
+      "learning_rate": 2.2655430294862962e-05,
+      "loss": 1.2949,
+      "step": 12071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.879801943531746,
+      "learning_rate": 2.2651278157455413e-05,
+      "loss": 0.793,
+      "step": 12072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4708424645738165,
+      "learning_rate": 2.264712608541184e-05,
+      "loss": 0.7212,
+      "step": 12073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3707560948095607,
+      "learning_rate": 2.2642974078847785e-05,
+      "loss": 0.9805,
+      "step": 12074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.467619238603491,
+      "learning_rate": 2.263882213787881e-05,
+      "loss": 1.2148,
+      "step": 12075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3008161713778605,
+      "learning_rate": 2.263467026262045e-05,
+      "loss": 0.9043,
+      "step": 12076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.150054560210899,
+      "learning_rate": 2.2630518453188253e-05,
+      "loss": 2.7715,
+      "step": 12077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.699404237891137,
+      "learning_rate": 2.262636670969775e-05,
+      "loss": 2.4199,
+      "step": 12078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1537234112852364,
+      "learning_rate": 2.2622215032264514e-05,
+      "loss": 1.1182,
+      "step": 12079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.95923257702558,
+      "learning_rate": 2.2618063421004054e-05,
+      "loss": 0.9238,
+      "step": 12080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.1118841047447265,
+      "learning_rate": 2.261391187603193e-05,
+      "loss": 1.373,
+      "step": 12081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3130728553351134,
+      "learning_rate": 2.260976039746366e-05,
+      "loss": 0.7762,
+      "step": 12082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2776850635182067,
+      "learning_rate": 2.2605608985414788e-05,
+      "loss": 0.8281,
+      "step": 12083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.401842415454848,
+      "learning_rate": 2.2601457640000842e-05,
+      "loss": 1.2207,
+      "step": 12084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6715324137608096,
+      "learning_rate": 2.2597306361337355e-05,
+      "loss": 0.8086,
+      "step": 12085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.884573380580161,
+      "learning_rate": 2.259315514953985e-05,
+      "loss": 0.5642,
+      "step": 12086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.589187377791193,
+      "learning_rate": 2.2589004004723863e-05,
+      "loss": 1.1934,
+      "step": 12087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.990650389751892,
+      "learning_rate": 2.2584852927004906e-05,
+      "loss": 0.7773,
+      "step": 12088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7917667389958245,
+      "learning_rate": 2.2580701916498507e-05,
+      "loss": 1.9062,
+      "step": 12089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5683261011636263,
+      "learning_rate": 2.25765509733202e-05,
+      "loss": 1.1421,
+      "step": 12090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.621502276242257,
+      "learning_rate": 2.2572400097585484e-05,
+      "loss": 1.1162,
+      "step": 12091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.370874039733451,
+      "learning_rate": 2.256824928940989e-05,
+      "loss": 0.6484,
+      "step": 12092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.04523081663727,
+      "learning_rate": 2.2564098548908925e-05,
+      "loss": 1.6289,
+      "step": 12093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.164719520824156,
+      "learning_rate": 2.255994787619811e-05,
+      "loss": 0.9043,
+      "step": 12094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5671449756408875,
+      "learning_rate": 2.2555797271392948e-05,
+      "loss": 0.9844,
+      "step": 12095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1907408548974137,
+      "learning_rate": 2.255164673460896e-05,
+      "loss": 1.1865,
+      "step": 12096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.483110549105503,
+      "learning_rate": 2.2547496265961626e-05,
+      "loss": 0.9536,
+      "step": 12097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.541143790904451,
+      "learning_rate": 2.254334586556649e-05,
+      "loss": 0.9102,
+      "step": 12098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9057312556614407,
+      "learning_rate": 2.2539195533539036e-05,
+      "loss": 0.9169,
+      "step": 12099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3974147616957633,
+      "learning_rate": 2.253504526999477e-05,
+      "loss": 1.4033,
+      "step": 12100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.123807348023774,
+      "learning_rate": 2.2530895075049187e-05,
+      "loss": 0.7979,
+      "step": 12101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.490839895667861,
+      "learning_rate": 2.2526744948817792e-05,
+      "loss": 1.5156,
+      "step": 12102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.470784186113663,
+      "learning_rate": 2.2522594891416078e-05,
+      "loss": 1.5703,
+      "step": 12103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4698057211019235,
+      "learning_rate": 2.251844490295954e-05,
+      "loss": 0.9502,
+      "step": 12104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.089185179154198,
+      "learning_rate": 2.2514294983563666e-05,
+      "loss": 1.2217,
+      "step": 12105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.6272104282497875,
+      "learning_rate": 2.2510145133343944e-05,
+      "loss": 1.3887,
+      "step": 12106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.878535040484525,
+      "learning_rate": 2.250599535241589e-05,
+      "loss": 1.0859,
+      "step": 12107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8542362241224772,
+      "learning_rate": 2.2501845640894952e-05,
+      "loss": 0.9365,
+      "step": 12108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.519775662654555,
+      "learning_rate": 2.2497695998896644e-05,
+      "loss": 1.0088,
+      "step": 12109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.033087890951372,
+      "learning_rate": 2.2493546426536434e-05,
+      "loss": 0.5022,
+      "step": 12110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.025776692954284,
+      "learning_rate": 2.2489396923929808e-05,
+      "loss": 1.1484,
+      "step": 12111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9167135136172753,
+      "learning_rate": 2.248524749119224e-05,
+      "loss": 1.0996,
+      "step": 12112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.072331276863301,
+      "learning_rate": 2.2481098128439213e-05,
+      "loss": 0.9766,
+      "step": 12113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.913883683551759,
+      "learning_rate": 2.2476948835786196e-05,
+      "loss": 1.3281,
+      "step": 12114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6528176964313372,
+      "learning_rate": 2.2472799613348673e-05,
+      "loss": 1.2021,
+      "step": 12115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.836053605897675,
+      "learning_rate": 2.246865046124209e-05,
+      "loss": 1.9785,
+      "step": 12116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9751200547557675,
+      "learning_rate": 2.246450137958195e-05,
+      "loss": 0.6396,
+      "step": 12117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5043260338501416,
+      "learning_rate": 2.24603523684837e-05,
+      "loss": 0.5991,
+      "step": 12118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4264634504570246,
+      "learning_rate": 2.2456203428062816e-05,
+      "loss": 1.2598,
+      "step": 12119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1046172562894165,
+      "learning_rate": 2.245205455843475e-05,
+      "loss": 0.8633,
+      "step": 12120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.56125947992579,
+      "learning_rate": 2.244790575971497e-05,
+      "loss": 1.4043,
+      "step": 12121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2537745051829594,
+      "learning_rate": 2.2443757032018932e-05,
+      "loss": 0.9561,
+      "step": 12122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3242383427154754,
+      "learning_rate": 2.24396083754621e-05,
+      "loss": 0.9106,
+      "step": 12123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8960270276175328,
+      "learning_rate": 2.243545979015992e-05,
+      "loss": 0.6165,
+      "step": 12124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.577036309847603,
+      "learning_rate": 2.243131127622784e-05,
+      "loss": 1.082,
+      "step": 12125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9581530520529387,
+      "learning_rate": 2.2427162833781334e-05,
+      "loss": 0.9072,
+      "step": 12126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.60261596365977,
+      "learning_rate": 2.242301446293584e-05,
+      "loss": 1.0986,
+      "step": 12127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0775719387117406,
+      "learning_rate": 2.24188661638068e-05,
+      "loss": 1.3496,
+      "step": 12128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.221690643599882,
+      "learning_rate": 2.2414717936509666e-05,
+      "loss": 0.9274,
+      "step": 12129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6978734330845846,
+      "learning_rate": 2.241056978115988e-05,
+      "loss": 0.6377,
+      "step": 12130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6461701301141747,
+      "learning_rate": 2.240642169787288e-05,
+      "loss": 0.6463,
+      "step": 12131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.321835810271107,
+      "learning_rate": 2.2402273686764112e-05,
+      "loss": 1.0645,
+      "step": 12132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2988930484805183,
+      "learning_rate": 2.2398125747949005e-05,
+      "loss": 1.1357,
+      "step": 12133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5445704418768007,
+      "learning_rate": 2.2393977881542997e-05,
+      "loss": 1.2598,
+      "step": 12134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.096089767695439,
+      "learning_rate": 2.2389830087661526e-05,
+      "loss": 1.0449,
+      "step": 12135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.078601780869182,
+      "learning_rate": 2.2385682366420025e-05,
+      "loss": 1.4277,
+      "step": 12136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.590952272070534,
+      "learning_rate": 2.2381534717933917e-05,
+      "loss": 0.7671,
+      "step": 12137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.04863497489118,
+      "learning_rate": 2.2377387142318633e-05,
+      "loss": 0.9131,
+      "step": 12138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3851612959407604,
+      "learning_rate": 2.2373239639689596e-05,
+      "loss": 1.3398,
+      "step": 12139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.51432985035891,
+      "learning_rate": 2.2369092210162234e-05,
+      "loss": 1.5703,
+      "step": 12140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.556722221500376,
+      "learning_rate": 2.236494485385196e-05,
+      "loss": 1.0869,
+      "step": 12141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0133230386348706,
+      "learning_rate": 2.2360797570874205e-05,
+      "loss": 1.1387,
+      "step": 12142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7974143567451213,
+      "learning_rate": 2.2356650361344366e-05,
+      "loss": 1.2031,
+      "step": 12143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.512161352072715,
+      "learning_rate": 2.2352503225377878e-05,
+      "loss": 1.2754,
+      "step": 12144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.090412905493375,
+      "learning_rate": 2.2348356163090154e-05,
+      "loss": 1.2061,
+      "step": 12145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8867279747526169,
+      "learning_rate": 2.234420917459659e-05,
+      "loss": 0.8975,
+      "step": 12146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5948472057775938,
+      "learning_rate": 2.234006226001261e-05,
+      "loss": 1.2188,
+      "step": 12147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7259733334116523,
+      "learning_rate": 2.233591541945361e-05,
+      "loss": 0.9121,
+      "step": 12148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.1329906786877295,
+      "learning_rate": 2.2331768653035004e-05,
+      "loss": 1.4648,
+      "step": 12149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.269110999152726,
+      "learning_rate": 2.2327621960872187e-05,
+      "loss": 1.3584,
+      "step": 12150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.105320743036745,
+      "learning_rate": 2.2323475343080564e-05,
+      "loss": 0.6113,
+      "step": 12151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.149561572625265,
+      "learning_rate": 2.231932879977553e-05,
+      "loss": 0.7256,
+      "step": 12152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.344851566847762,
+      "learning_rate": 2.2315182331072477e-05,
+      "loss": 0.8906,
+      "step": 12153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6071921715299604,
+      "learning_rate": 2.231103593708682e-05,
+      "loss": 1.375,
+      "step": 12154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5506982483145824,
+      "learning_rate": 2.2306889617933928e-05,
+      "loss": 0.6201,
+      "step": 12155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.259441804228008,
+      "learning_rate": 2.2302743373729205e-05,
+      "loss": 0.6733,
+      "step": 12156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.729996605428715,
+      "learning_rate": 2.2298597204588044e-05,
+      "loss": 1.7344,
+      "step": 12157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2483764724985362,
+      "learning_rate": 2.2294451110625814e-05,
+      "loss": 1.2969,
+      "step": 12158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5305736250313093,
+      "learning_rate": 2.2290305091957914e-05,
+      "loss": 0.9961,
+      "step": 12159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.107178075414319,
+      "learning_rate": 2.228615914869972e-05,
+      "loss": 1.1318,
+      "step": 12160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.10997716572116,
+      "learning_rate": 2.228201328096661e-05,
+      "loss": 1.9727,
+      "step": 12161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.759487864345125,
+      "learning_rate": 2.2277867488873955e-05,
+      "loss": 0.9102,
+      "step": 12162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6228283573118056,
+      "learning_rate": 2.2273721772537146e-05,
+      "loss": 1.0703,
+      "step": 12163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.445123699313918,
+      "learning_rate": 2.2269576132071558e-05,
+      "loss": 1.1643,
+      "step": 12164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.0336258295894,
+      "learning_rate": 2.2265430567592544e-05,
+      "loss": 1.1543,
+      "step": 12165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.825359281927405,
+      "learning_rate": 2.2261285079215492e-05,
+      "loss": 1.2793,
+      "step": 12166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.538972705812185,
+      "learning_rate": 2.2257139667055757e-05,
+      "loss": 1.4531,
+      "step": 12167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.240580231925606,
+      "learning_rate": 2.2252994331228713e-05,
+      "loss": 0.875,
+      "step": 12168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.039876062739416,
+      "learning_rate": 2.2248849071849712e-05,
+      "loss": 1.5332,
+      "step": 12169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.645176116115246,
+      "learning_rate": 2.224470388903413e-05,
+      "loss": 1.0146,
+      "step": 12170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3889159033477556,
+      "learning_rate": 2.2240558782897303e-05,
+      "loss": 1.2266,
+      "step": 12171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7317374264095173,
+      "learning_rate": 2.2236413753554606e-05,
+      "loss": 1.0029,
+      "step": 12172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9505836126990315,
+      "learning_rate": 2.2232268801121396e-05,
+      "loss": 0.9277,
+      "step": 12173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.979637645116375,
+      "learning_rate": 2.2228123925713016e-05,
+      "loss": 1.1084,
+      "step": 12174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3482873381865033,
+      "learning_rate": 2.2223979127444816e-05,
+      "loss": 0.9443,
+      "step": 12175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.863350369073293,
+      "learning_rate": 2.221983440643215e-05,
+      "loss": 1.5977,
+      "step": 12176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.316487916765382,
+      "learning_rate": 2.2215689762790362e-05,
+      "loss": 0.9336,
+      "step": 12177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.047148352857733,
+      "learning_rate": 2.2211545196634794e-05,
+      "loss": 0.7388,
+      "step": 12178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.136652692648926,
+      "learning_rate": 2.2207400708080787e-05,
+      "loss": 0.6731,
+      "step": 12179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.618378100050321,
+      "learning_rate": 2.2203256297243673e-05,
+      "loss": 1.4512,
+      "step": 12180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1298393794287698,
+      "learning_rate": 2.219911196423881e-05,
+      "loss": 0.8545,
+      "step": 12181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4557080856328,
+      "learning_rate": 2.2194967709181513e-05,
+      "loss": 0.8027,
+      "step": 12182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.620471772532272,
+      "learning_rate": 2.219082353218713e-05,
+      "loss": 0.3361,
+      "step": 12183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.209911422708711,
+      "learning_rate": 2.218667943337098e-05,
+      "loss": 1.3682,
+      "step": 12184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.673006118085897,
+      "learning_rate": 2.2182535412848397e-05,
+      "loss": 1.5664,
+      "step": 12185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.157089733310534,
+      "learning_rate": 2.2178391470734704e-05,
+      "loss": 0.8008,
+      "step": 12186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4546403031695574,
+      "learning_rate": 2.217424760714523e-05,
+      "loss": 0.8975,
+      "step": 12187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.332440057701852,
+      "learning_rate": 2.2170103822195296e-05,
+      "loss": 1.2881,
+      "step": 12188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5710443538903855,
+      "learning_rate": 2.2165960116000222e-05,
+      "loss": 0.874,
+      "step": 12189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.244154861179279,
+      "learning_rate": 2.2161816488675312e-05,
+      "loss": 1.3887,
+      "step": 12190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8615370289954065,
+      "learning_rate": 2.2157672940335903e-05,
+      "loss": 0.6006,
+      "step": 12191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1900114749235686,
+      "learning_rate": 2.2153529471097303e-05,
+      "loss": 0.9893,
+      "step": 12192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.668163745316064,
+      "learning_rate": 2.2149386081074812e-05,
+      "loss": 1.5288,
+      "step": 12193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0993772582450547,
+      "learning_rate": 2.2145242770383754e-05,
+      "loss": 0.8975,
+      "step": 12194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.69512761905547,
+      "learning_rate": 2.214109953913942e-05,
+      "loss": 1.2715,
+      "step": 12195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.132824629267464,
+      "learning_rate": 2.2136956387457124e-05,
+      "loss": 1.0464,
+      "step": 12196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.071781764569566,
+      "learning_rate": 2.2132813315452168e-05,
+      "loss": 1.0547,
+      "step": 12197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.198461805479037,
+      "learning_rate": 2.2128670323239846e-05,
+      "loss": 0.8403,
+      "step": 12198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9978845486022525,
+      "learning_rate": 2.2124527410935455e-05,
+      "loss": 1.0781,
+      "step": 12199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.612265210764638,
+      "learning_rate": 2.2120384578654307e-05,
+      "loss": 1.0293,
+      "step": 12200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078649583732059,
+      "learning_rate": 2.2116241826511676e-05,
+      "loss": 1.0977,
+      "step": 12201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.937335557595945,
+      "learning_rate": 2.211209915462287e-05,
+      "loss": 0.7607,
+      "step": 12202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.585850740112503,
+      "learning_rate": 2.2107956563103162e-05,
+      "loss": 1.1348,
+      "step": 12203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7725300768501278,
+      "learning_rate": 2.2103814052067852e-05,
+      "loss": 0.8881,
+      "step": 12204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.923258989375706,
+      "learning_rate": 2.2099671621632213e-05,
+      "loss": 0.9902,
+      "step": 12205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.752034366467862,
+      "learning_rate": 2.2095529271911533e-05,
+      "loss": 1.0405,
+      "step": 12206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.409839655930318,
+      "learning_rate": 2.209138700302109e-05,
+      "loss": 0.6641,
+      "step": 12207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.287609589666713,
+      "learning_rate": 2.2087244815076154e-05,
+      "loss": 1.2695,
+      "step": 12208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2798054157561576,
+      "learning_rate": 2.208310270819202e-05,
+      "loss": 0.3818,
+      "step": 12209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.450848634843512,
+      "learning_rate": 2.207896068248395e-05,
+      "loss": 1.3047,
+      "step": 12210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.801118632919343,
+      "learning_rate": 2.2074818738067217e-05,
+      "loss": 0.9243,
+      "step": 12211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.644439054685768,
+      "learning_rate": 2.207067687505708e-05,
+      "loss": 1.5391,
+      "step": 12212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2428245569834933,
+      "learning_rate": 2.206653509356882e-05,
+      "loss": 1.1943,
+      "step": 12213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6545619312965036,
+      "learning_rate": 2.206239339371769e-05,
+      "loss": 0.6465,
+      "step": 12214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6520257009121044,
+      "learning_rate": 2.2058251775618954e-05,
+      "loss": 0.9512,
+      "step": 12215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.335316326563402,
+      "learning_rate": 2.2054110239387883e-05,
+      "loss": 1.2305,
+      "step": 12216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7083936618604296,
+      "learning_rate": 2.2049968785139707e-05,
+      "loss": 0.6123,
+      "step": 12217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5901059458621343,
+      "learning_rate": 2.204582741298971e-05,
+      "loss": 0.9873,
+      "step": 12218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9382928641116233,
+      "learning_rate": 2.2041686123053136e-05,
+      "loss": 1.0488,
+      "step": 12219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.500843192003447,
+      "learning_rate": 2.203754491544523e-05,
+      "loss": 0.8931,
+      "step": 12220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3362896842783116,
+      "learning_rate": 2.2033403790281245e-05,
+      "loss": 1.0752,
+      "step": 12221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.326022541567135,
+      "learning_rate": 2.202926274767642e-05,
+      "loss": 1.6699,
+      "step": 12222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.222376478987183,
+      "learning_rate": 2.2025121787746012e-05,
+      "loss": 1.2539,
+      "step": 12223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.775104593143823,
+      "learning_rate": 2.2020980910605247e-05,
+      "loss": 1.5059,
+      "step": 12224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.80449035259081,
+      "learning_rate": 2.2016840116369376e-05,
+      "loss": 1.4883,
+      "step": 12225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7727459575946636,
+      "learning_rate": 2.2012699405153623e-05,
+      "loss": 1.7285,
+      "step": 12226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.207964865287627,
+      "learning_rate": 2.2008558777073223e-05,
+      "loss": 1.1738,
+      "step": 12227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.07337569580876,
+      "learning_rate": 2.2004418232243425e-05,
+      "loss": 0.6846,
+      "step": 12228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.363328679031054,
+      "learning_rate": 2.2000277770779447e-05,
+      "loss": 0.6538,
+      "step": 12229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.741994823848811,
+      "learning_rate": 2.199613739279652e-05,
+      "loss": 1.209,
+      "step": 12230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3758950636044287,
+      "learning_rate": 2.199199709840986e-05,
+      "loss": 1.002,
+      "step": 12231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.363957318310758,
+      "learning_rate": 2.1987856887734703e-05,
+      "loss": 1.2881,
+      "step": 12232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.412474668695945,
+      "learning_rate": 2.198371676088626e-05,
+      "loss": 1.207,
+      "step": 12233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.746903665140881,
+      "learning_rate": 2.1979576717979756e-05,
+      "loss": 1.1973,
+      "step": 12234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9294636708536819,
+      "learning_rate": 2.1975436759130394e-05,
+      "loss": 0.4985,
+      "step": 12235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.216990758615204,
+      "learning_rate": 2.197129688445339e-05,
+      "loss": 0.8047,
+      "step": 12236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7867690132253566,
+      "learning_rate": 2.196715709406397e-05,
+      "loss": 1.1543,
+      "step": 12237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.0672957699030885,
+      "learning_rate": 2.1963017388077333e-05,
+      "loss": 1.2832,
+      "step": 12238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0518419411000752,
+      "learning_rate": 2.1958877766608685e-05,
+      "loss": 0.825,
+      "step": 12239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.263830686731153,
+      "learning_rate": 2.1954738229773237e-05,
+      "loss": 1.1104,
+      "step": 12240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.203242954857099,
+      "learning_rate": 2.1950598777686178e-05,
+      "loss": 1.7305,
+      "step": 12241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.443456309788448,
+      "learning_rate": 2.1946459410462715e-05,
+      "loss": 1.124,
+      "step": 12242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.198357124498397,
+      "learning_rate": 2.194232012821804e-05,
+      "loss": 0.9648,
+      "step": 12243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.064675657464417,
+      "learning_rate": 2.193818093106736e-05,
+      "loss": 1.582,
+      "step": 12244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.508030358406872,
+      "learning_rate": 2.1934041819125843e-05,
+      "loss": 1.041,
+      "step": 12245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.0475588187643945,
+      "learning_rate": 2.19299027925087e-05,
+      "loss": 1.2266,
+      "step": 12246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4195923877991437,
+      "learning_rate": 2.1925763851331123e-05,
+      "loss": 1.3799,
+      "step": 12247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9872773015606797,
+      "learning_rate": 2.1921624995708276e-05,
+      "loss": 0.9111,
+      "step": 12248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1262171091746476,
+      "learning_rate": 2.1917486225755358e-05,
+      "loss": 0.6816,
+      "step": 12249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6006693638632186,
+      "learning_rate": 2.1913347541587538e-05,
+      "loss": 0.8906,
+      "step": 12250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.537964100941843,
+      "learning_rate": 2.1909208943320006e-05,
+      "loss": 1.248,
+      "step": 12251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9232248667516862,
+      "learning_rate": 2.1905070431067924e-05,
+      "loss": 1.4648,
+      "step": 12252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.952671076689983,
+      "learning_rate": 2.1900932004946483e-05,
+      "loss": 0.9326,
+      "step": 12253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5790707866599747,
+      "learning_rate": 2.189679366507083e-05,
+      "loss": 1.1377,
+      "step": 12254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.126971119072992,
+      "learning_rate": 2.1892655411556143e-05,
+      "loss": 0.9111,
+      "step": 12255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.804161076349114,
+      "learning_rate": 2.1888517244517594e-05,
+      "loss": 1.1953,
+      "step": 12256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.032367775496023,
+      "learning_rate": 2.1884379164070352e-05,
+      "loss": 0.9395,
+      "step": 12257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8121185043998564,
+      "learning_rate": 2.1880241170329567e-05,
+      "loss": 0.5986,
+      "step": 12258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.213178784839597,
+      "learning_rate": 2.18761032634104e-05,
+      "loss": 1.1152,
+      "step": 12259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.865781234947497,
+      "learning_rate": 2.1871965443428007e-05,
+      "loss": 0.9951,
+      "step": 12260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.752635733075466,
+      "learning_rate": 2.186782771049754e-05,
+      "loss": 0.8135,
+      "step": 12261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.486024613912439,
+      "learning_rate": 2.1863690064734156e-05,
+      "loss": 1.5176,
+      "step": 12262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.401831297763492,
+      "learning_rate": 2.1859552506253e-05,
+      "loss": 1.002,
+      "step": 12263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0798131066501506,
+      "learning_rate": 2.185541503516921e-05,
+      "loss": 1.5195,
+      "step": 12264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7611857282446275,
+      "learning_rate": 2.1851277651597946e-05,
+      "loss": 1.5547,
+      "step": 12265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8323832592288674,
+      "learning_rate": 2.184714035565435e-05,
+      "loss": 1.249,
+      "step": 12266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.9705753431288775,
+      "learning_rate": 2.1843003147453544e-05,
+      "loss": 1.2559,
+      "step": 12267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2657653765105605,
+      "learning_rate": 2.1838866027110684e-05,
+      "loss": 0.8777,
+      "step": 12268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.206548731162498,
+      "learning_rate": 2.183472899474089e-05,
+      "loss": 1.1738,
+      "step": 12269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.36221625599448,
+      "learning_rate": 2.1830592050459303e-05,
+      "loss": 0.875,
+      "step": 12270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2025919600826773,
+      "learning_rate": 2.1826455194381046e-05,
+      "loss": 0.8472,
+      "step": 12271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.226358076381973,
+      "learning_rate": 2.1822318426621255e-05,
+      "loss": 0.8267,
+      "step": 12272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.032517344844856,
+      "learning_rate": 2.181818174729503e-05,
+      "loss": 0.6543,
+      "step": 12273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3188385687610062,
+      "learning_rate": 2.1814045156517528e-05,
+      "loss": 1.1387,
+      "step": 12274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.510563277610471,
+      "learning_rate": 2.180990865440385e-05,
+      "loss": 1.0674,
+      "step": 12275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5909772201911743,
+      "learning_rate": 2.1805772241069118e-05,
+      "loss": 0.7305,
+      "step": 12276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.030661133331966,
+      "learning_rate": 2.1801635916628437e-05,
+      "loss": 0.9658,
+      "step": 12277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.37220673046023,
+      "learning_rate": 2.1797499681196937e-05,
+      "loss": 0.6377,
+      "step": 12278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8962057677046977,
+      "learning_rate": 2.1793363534889708e-05,
+      "loss": 1.0645,
+      "step": 12279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.768019177993506,
+      "learning_rate": 2.1789227477821872e-05,
+      "loss": 0.688,
+      "step": 12280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.452037802824936,
+      "learning_rate": 2.1785091510108524e-05,
+      "loss": 1.3545,
+      "step": 12281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.899918863052934,
+      "learning_rate": 2.1780955631864763e-05,
+      "loss": 1.0283,
+      "step": 12282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.861344302215417,
+      "learning_rate": 2.177681984320571e-05,
+      "loss": 1.3438,
+      "step": 12283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8382168774338137,
+      "learning_rate": 2.177268414424644e-05,
+      "loss": 0.9775,
+      "step": 12284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.930620131501791,
+      "learning_rate": 2.1768548535102065e-05,
+      "loss": 0.9434,
+      "step": 12285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.324717147486263,
+      "learning_rate": 2.176441301588766e-05,
+      "loss": 1.1133,
+      "step": 12286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.045868066536625,
+      "learning_rate": 2.176027758671833e-05,
+      "loss": 1.2578,
+      "step": 12287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.025981938152634,
+      "learning_rate": 2.1756142247709156e-05,
+      "loss": 0.5388,
+      "step": 12288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.073971935299074,
+      "learning_rate": 2.1752006998975223e-05,
+      "loss": 1.2949,
+      "step": 12289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0042456623880756,
+      "learning_rate": 2.1747871840631607e-05,
+      "loss": 0.5981,
+      "step": 12290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.270506447412224,
+      "learning_rate": 2.1743736772793398e-05,
+      "loss": 1.2891,
+      "step": 12291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.827695612278535,
+      "learning_rate": 2.1739601795575655e-05,
+      "loss": 1.0347,
+      "step": 12292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.498375705454888,
+      "learning_rate": 2.1735466909093482e-05,
+      "loss": 0.7402,
+      "step": 12293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.30134111826177,
+      "learning_rate": 2.1731332113461934e-05,
+      "loss": 1.0908,
+      "step": 12294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0126728054752006,
+      "learning_rate": 2.1727197408796084e-05,
+      "loss": 0.8301,
+      "step": 12295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8077715959570813,
+      "learning_rate": 2.1723062795210995e-05,
+      "loss": 1.2324,
+      "step": 12296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.025649896609321,
+      "learning_rate": 2.1718928272821736e-05,
+      "loss": 1.4883,
+      "step": 12297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2154971711337197,
+      "learning_rate": 2.1714793841743367e-05,
+      "loss": 1.1445,
+      "step": 12298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2607930771523046,
+      "learning_rate": 2.171065950209095e-05,
+      "loss": 0.9033,
+      "step": 12299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.015013727751889,
+      "learning_rate": 2.1706525253979534e-05,
+      "loss": 0.6167,
+      "step": 12300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.520240404943242,
+      "learning_rate": 2.1702391097524178e-05,
+      "loss": 1.2812,
+      "step": 12301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.768557270997472,
+      "learning_rate": 2.1698257032839943e-05,
+      "loss": 0.7437,
+      "step": 12302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.361643606272935,
+      "learning_rate": 2.1694123060041865e-05,
+      "loss": 0.6938,
+      "step": 12303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9305206230107337,
+      "learning_rate": 2.1689989179245003e-05,
+      "loss": 1.3184,
+      "step": 12304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2018955755605822,
+      "learning_rate": 2.1685855390564387e-05,
+      "loss": 0.6377,
+      "step": 12305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5279413437324743,
+      "learning_rate": 2.1681721694115077e-05,
+      "loss": 1.4941,
+      "step": 12306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.999803729973491,
+      "learning_rate": 2.1677588090012092e-05,
+      "loss": 0.8848,
+      "step": 12307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.921004121294421,
+      "learning_rate": 2.1673454578370484e-05,
+      "loss": 0.9082,
+      "step": 12308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.945769014065515,
+      "learning_rate": 2.166932115930527e-05,
+      "loss": 1.4238,
+      "step": 12309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.145712744878863,
+      "learning_rate": 2.1665187832931494e-05,
+      "loss": 1.377,
+      "step": 12310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1070624718501385,
+      "learning_rate": 2.1661054599364196e-05,
+      "loss": 1.3633,
+      "step": 12311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.796279843903965,
+      "learning_rate": 2.1656921458718377e-05,
+      "loss": 1.0479,
+      "step": 12312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7623284299572255,
+      "learning_rate": 2.1652788411109075e-05,
+      "loss": 0.6035,
+      "step": 12313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.996893556416097,
+      "learning_rate": 2.1648655456651313e-05,
+      "loss": 0.6462,
+      "step": 12314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.057921906200885,
+      "learning_rate": 2.1644522595460102e-05,
+      "loss": 0.833,
+      "step": 12315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.602606658003301,
+      "learning_rate": 2.164038982765047e-05,
+      "loss": 2.7754,
+      "step": 12316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.725065731592389,
+      "learning_rate": 2.163625715333741e-05,
+      "loss": 1.1875,
+      "step": 12317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0736009500295323,
+      "learning_rate": 2.1632124572635955e-05,
+      "loss": 1.375,
+      "step": 12318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9087499266538874,
+      "learning_rate": 2.1627992085661087e-05,
+      "loss": 0.7656,
+      "step": 12319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.389302010280356,
+      "learning_rate": 2.162385969252783e-05,
+      "loss": 1.1826,
+      "step": 12320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.5764546572887586,
+      "learning_rate": 2.1619727393351194e-05,
+      "loss": 0.7559,
+      "step": 12321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.186510106512947,
+      "learning_rate": 2.161559518824616e-05,
+      "loss": 0.6333,
+      "step": 12322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.632164215959527,
+      "learning_rate": 2.161146307732774e-05,
+      "loss": 0.5693,
+      "step": 12323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.81398813951821,
+      "learning_rate": 2.1607331060710918e-05,
+      "loss": 1.5586,
+      "step": 12324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1297073005586977,
+      "learning_rate": 2.1603199138510697e-05,
+      "loss": 1.168,
+      "step": 12325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3713755809326167,
+      "learning_rate": 2.1599067310842055e-05,
+      "loss": 0.8784,
+      "step": 12326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.561682222837145,
+      "learning_rate": 2.1594935577819993e-05,
+      "loss": 0.7695,
+      "step": 12327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.721850589331779,
+      "learning_rate": 2.1590803939559477e-05,
+      "loss": 0.8428,
+      "step": 12328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.751755674167625,
+      "learning_rate": 2.15866723961755e-05,
+      "loss": 0.6812,
+      "step": 12329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.474559970137799,
+      "learning_rate": 2.158254094778305e-05,
+      "loss": 1.0645,
+      "step": 12330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.618412361455613,
+      "learning_rate": 2.1578409594497085e-05,
+      "loss": 1.6426,
+      "step": 12331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0129204873509576,
+      "learning_rate": 2.157427833643259e-05,
+      "loss": 1.2578,
+      "step": 12332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6541584531775744,
+      "learning_rate": 2.157014717370454e-05,
+      "loss": 0.8276,
+      "step": 12333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.407582892301069,
+      "learning_rate": 2.1566016106427894e-05,
+      "loss": 1.1699,
+      "step": 12334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1010233247706864,
+      "learning_rate": 2.1561885134717626e-05,
+      "loss": 1.126,
+      "step": 12335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9782749966837483,
+      "learning_rate": 2.155775425868869e-05,
+      "loss": 1.3711,
+      "step": 12336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5441534549393037,
+      "learning_rate": 2.155362347845606e-05,
+      "loss": 0.8325,
+      "step": 12337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8310017996957137,
+      "learning_rate": 2.1549492794134668e-05,
+      "loss": 1.4238,
+      "step": 12338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9886115396512345,
+      "learning_rate": 2.1545362205839497e-05,
+      "loss": 1.0098,
+      "step": 12339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3537998645137734,
+      "learning_rate": 2.1541231713685493e-05,
+      "loss": 0.7832,
+      "step": 12340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6387608736899306,
+      "learning_rate": 2.15371013177876e-05,
+      "loss": 1.293,
+      "step": 12341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.919039922425194,
+      "learning_rate": 2.153297101826077e-05,
+      "loss": 1.1064,
+      "step": 12342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.391086110043829,
+      "learning_rate": 2.152884081521994e-05,
+      "loss": 0.854,
+      "step": 12343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0201111778030474,
+      "learning_rate": 2.152471070878006e-05,
+      "loss": 0.8154,
+      "step": 12344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3913541193068375,
+      "learning_rate": 2.1520580699056065e-05,
+      "loss": 1.1279,
+      "step": 12345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6592584066685774,
+      "learning_rate": 2.15164507861629e-05,
+      "loss": 0.9309,
+      "step": 12346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1583484997701277,
+      "learning_rate": 2.1512320970215474e-05,
+      "loss": 0.7592,
+      "step": 12347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.186700511337055,
+      "learning_rate": 2.1508191251328744e-05,
+      "loss": 0.957,
+      "step": 12348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7851053555337444,
+      "learning_rate": 2.1504061629617635e-05,
+      "loss": 1.3281,
+      "step": 12349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.182256900143211,
+      "learning_rate": 2.1499932105197062e-05,
+      "loss": 1.1289,
+      "step": 12350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.297814616725148,
+      "learning_rate": 2.149580267818196e-05,
+      "loss": 0.905,
+      "step": 12351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.918495592374961,
+      "learning_rate": 2.1491673348687238e-05,
+      "loss": 1.0352,
+      "step": 12352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5370775181133625,
+      "learning_rate": 2.1487544116827816e-05,
+      "loss": 0.356,
+      "step": 12353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.778426195733028,
+      "learning_rate": 2.1483414982718623e-05,
+      "loss": 1.1133,
+      "step": 12354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.846486802214369,
+      "learning_rate": 2.147928594647455e-05,
+      "loss": 0.7061,
+      "step": 12355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1730573989583157,
+      "learning_rate": 2.1475157008210507e-05,
+      "loss": 1.3398,
+      "step": 12356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2002044149279087,
+      "learning_rate": 2.1471028168041425e-05,
+      "loss": 1.0508,
+      "step": 12357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9243461453026613,
+      "learning_rate": 2.1466899426082184e-05,
+      "loss": 1.3184,
+      "step": 12358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9433989038072284,
+      "learning_rate": 2.14627707824477e-05,
+      "loss": 0.942,
+      "step": 12359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.461858170030483,
+      "learning_rate": 2.145864223725286e-05,
+      "loss": 0.9097,
+      "step": 12360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7344046237924393,
+      "learning_rate": 2.1454513790612572e-05,
+      "loss": 0.5186,
+      "step": 12361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4206023779613997,
+      "learning_rate": 2.145038544264171e-05,
+      "loss": 1.4883,
+      "step": 12362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3692633965767502,
+      "learning_rate": 2.1446257193455187e-05,
+      "loss": 1.1152,
+      "step": 12363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.545316579327624,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 1.3516,
+      "step": 12364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.819471492752428,
+      "learning_rate": 2.143800099189467e-05,
+      "loss": 1.1562,
+      "step": 12365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.568005114588888,
+      "learning_rate": 2.1433873039750434e-05,
+      "loss": 1.0342,
+      "step": 12366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.33842330893646,
+      "learning_rate": 2.1429745186850063e-05,
+      "loss": 0.9702,
+      "step": 12367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4647888080742746,
+      "learning_rate": 2.1425617433308436e-05,
+      "loss": 1.1387,
+      "step": 12368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7134280678272016,
+      "learning_rate": 2.1421489779240416e-05,
+      "loss": 1.0928,
+      "step": 12369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.153765465346559,
+      "learning_rate": 2.1417362224760882e-05,
+      "loss": 0.9114,
+      "step": 12370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.142184580370728,
+      "learning_rate": 2.1413234769984696e-05,
+      "loss": 1.2202,
+      "step": 12371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.191996636270571,
+      "learning_rate": 2.140910741502673e-05,
+      "loss": 1.2266,
+      "step": 12372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140302090758071,
+      "learning_rate": 2.140498016000184e-05,
+      "loss": 0.687,
+      "step": 12373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0640588129373056,
+      "learning_rate": 2.140085300502489e-05,
+      "loss": 1.1641,
+      "step": 12374
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.493347897737957,
+      "learning_rate": 2.1396725950210728e-05,
+      "loss": 1.6206,
+      "step": 12375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.996123193151042,
+      "learning_rate": 2.1392598995674226e-05,
+      "loss": 1.0332,
+      "step": 12376
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6769291775226645,
+      "learning_rate": 2.138847214153022e-05,
+      "loss": 0.9536,
+      "step": 12377
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9972623413656585,
+      "learning_rate": 2.138434538789357e-05,
+      "loss": 0.7266,
+      "step": 12378
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5070722516799417,
+      "learning_rate": 2.138021873487911e-05,
+      "loss": 1.5762,
+      "step": 12379
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5758159854901659,
+      "learning_rate": 2.1376092182601694e-05,
+      "loss": 0.3972,
+      "step": 12380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.959886832522543,
+      "learning_rate": 2.1371965731176152e-05,
+      "loss": 1.4062,
+      "step": 12381
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8548142378411536,
+      "learning_rate": 2.1367839380717334e-05,
+      "loss": 1.4707,
+      "step": 12382
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6443121634807283,
+      "learning_rate": 2.136371313134006e-05,
+      "loss": 1.4219,
+      "step": 12383
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0872124850809266,
+      "learning_rate": 2.1359586983159165e-05,
+      "loss": 0.8848,
+      "step": 12384
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6992259046180958,
+      "learning_rate": 2.135546093628949e-05,
+      "loss": 1.2402,
+      "step": 12385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.723653506615565,
+      "learning_rate": 2.135133499084585e-05,
+      "loss": 1.2891,
+      "step": 12386
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.250341104066717,
+      "learning_rate": 2.134720914694308e-05,
+      "loss": 0.9893,
+      "step": 12387
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6666100813902047,
+      "learning_rate": 2.1343083404695983e-05,
+      "loss": 1.543,
+      "step": 12388
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2646419102293818,
+      "learning_rate": 2.1338957764219396e-05,
+      "loss": 0.9102,
+      "step": 12389
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2710813917644606,
+      "learning_rate": 2.1334832225628115e-05,
+      "loss": 0.7983,
+      "step": 12390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4062009533924407,
+      "learning_rate": 2.1330706789036966e-05,
+      "loss": 1.1719,
+      "step": 12391
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.288289118499616,
+      "learning_rate": 2.132658145456075e-05,
+      "loss": 0.9844,
+      "step": 12392
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6339559816869293,
+      "learning_rate": 2.1322456222314274e-05,
+      "loss": 0.6045,
+      "step": 12393
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7967383495172373,
+      "learning_rate": 2.1318331092412342e-05,
+      "loss": 0.8831,
+      "step": 12394
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0600964203148133,
+      "learning_rate": 2.1314206064969765e-05,
+      "loss": 1.0176,
+      "step": 12395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.254889579488547,
+      "learning_rate": 2.131008114010133e-05,
+      "loss": 0.8467,
+      "step": 12396
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1348856816295125,
+      "learning_rate": 2.1305956317921838e-05,
+      "loss": 1.0996,
+      "step": 12397
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.35967144827435,
+      "learning_rate": 2.1301831598546072e-05,
+      "loss": 1.209,
+      "step": 12398
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5621565598076845,
+      "learning_rate": 2.1297706982088835e-05,
+      "loss": 0.8372,
+      "step": 12399
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.939730994185738,
+      "learning_rate": 2.1293582468664898e-05,
+      "loss": 0.9326,
+      "step": 12400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.001120613862721,
+      "learning_rate": 2.1289458058389063e-05,
+      "loss": 1.04,
+      "step": 12401
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4501723702534703,
+      "learning_rate": 2.1285333751376085e-05,
+      "loss": 1.1577,
+      "step": 12402
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.813935146807227,
+      "learning_rate": 2.1281209547740757e-05,
+      "loss": 1.3008,
+      "step": 12403
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.44329326873902,
+      "learning_rate": 2.127708544759786e-05,
+      "loss": 1.0879,
+      "step": 12404
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9289414740250717,
+      "learning_rate": 2.127296145106216e-05,
+      "loss": 2.0508,
+      "step": 12405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7636738591810457,
+      "learning_rate": 2.1268837558248428e-05,
+      "loss": 1.1475,
+      "step": 12406
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.785671203039103,
+      "learning_rate": 2.1264713769271422e-05,
+      "loss": 1.0752,
+      "step": 12407
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2754875457161328,
+      "learning_rate": 2.1260590084245914e-05,
+      "loss": 0.8643,
+      "step": 12408
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.465104533018254,
+      "learning_rate": 2.1256466503286658e-05,
+      "loss": 1.21,
+      "step": 12409
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.135305305474122,
+      "learning_rate": 2.1252343026508422e-05,
+      "loss": 1.2705,
+      "step": 12410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.540654736634573,
+      "learning_rate": 2.124821965402595e-05,
+      "loss": 0.8838,
+      "step": 12411
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.077272991173839,
+      "learning_rate": 2.124409638595399e-05,
+      "loss": 1.0771,
+      "step": 12412
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8953119993338072,
+      "learning_rate": 2.1239973222407304e-05,
+      "loss": 1.1816,
+      "step": 12413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2701242955666054,
+      "learning_rate": 2.1235850163500636e-05,
+      "loss": 1.375,
+      "step": 12414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6052252385443726,
+      "learning_rate": 2.1231727209348722e-05,
+      "loss": 1.5762,
+      "step": 12415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0607186919257554,
+      "learning_rate": 2.1227604360066314e-05,
+      "loss": 0.4878,
+      "step": 12416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2151489047595803,
+      "learning_rate": 2.122348161576813e-05,
+      "loss": 0.8101,
+      "step": 12417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.259677736886138,
+      "learning_rate": 2.1219358976568927e-05,
+      "loss": 0.8955,
+      "step": 12418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9636743300031587,
+      "learning_rate": 2.1215236442583414e-05,
+      "loss": 1.1328,
+      "step": 12419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.285628728031647,
+      "learning_rate": 2.1211114013926336e-05,
+      "loss": 0.5215,
+      "step": 12420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6412731932575433,
+      "learning_rate": 2.1206991690712404e-05,
+      "loss": 1.3633,
+      "step": 12421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.264878204145038,
+      "learning_rate": 2.1202869473056353e-05,
+      "loss": 0.7695,
+      "step": 12422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4411713848009304,
+      "learning_rate": 2.1198747361072903e-05,
+      "loss": 1.127,
+      "step": 12423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6731022029793854,
+      "learning_rate": 2.1194625354876766e-05,
+      "loss": 1.0488,
+      "step": 12424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.351142672356597,
+      "learning_rate": 2.1190503454582655e-05,
+      "loss": 1.3271,
+      "step": 12425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.534438303428493,
+      "learning_rate": 2.1186381660305282e-05,
+      "loss": 1.0,
+      "step": 12426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.32355491783338,
+      "learning_rate": 2.1182259972159358e-05,
+      "loss": 1.4648,
+      "step": 12427
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9680240443125459,
+      "learning_rate": 2.117813839025958e-05,
+      "loss": 1.0049,
+      "step": 12428
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9453956555590668,
+      "learning_rate": 2.117401691472066e-05,
+      "loss": 0.9658,
+      "step": 12429
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9875667464846476,
+      "learning_rate": 2.1169895545657285e-05,
+      "loss": 0.4039,
+      "step": 12430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2295481156407626,
+      "learning_rate": 2.1165774283184153e-05,
+      "loss": 1.2832,
+      "step": 12431
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4356809484746855,
+      "learning_rate": 2.1161653127415967e-05,
+      "loss": 1.1025,
+      "step": 12432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9744852983191885,
+      "learning_rate": 2.1157532078467417e-05,
+      "loss": 0.9707,
+      "step": 12433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.262399100061208,
+      "learning_rate": 2.115341113645318e-05,
+      "loss": 1.1562,
+      "step": 12434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.980758162319711,
+      "learning_rate": 2.114929030148795e-05,
+      "loss": 1.1318,
+      "step": 12435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7933642127424982,
+      "learning_rate": 2.11451695736864e-05,
+      "loss": 0.6177,
+      "step": 12436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.634327585516077,
+      "learning_rate": 2.1141048953163215e-05,
+      "loss": 0.7822,
+      "step": 12437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1974667876666416,
+      "learning_rate": 2.1136928440033065e-05,
+      "loss": 0.9629,
+      "step": 12438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0538064998438865,
+      "learning_rate": 2.1132808034410623e-05,
+      "loss": 1.1118,
+      "step": 12439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1385956086136964,
+      "learning_rate": 2.112868773641055e-05,
+      "loss": 1.2607,
+      "step": 12440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6737857701003027,
+      "learning_rate": 2.112456754614753e-05,
+      "loss": 0.6221,
+      "step": 12441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.175552404715032,
+      "learning_rate": 2.112044746373622e-05,
+      "loss": 1.4131,
+      "step": 12442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.5559229271892825,
+      "learning_rate": 2.1116327489291272e-05,
+      "loss": 1.5691,
+      "step": 12443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.70950229317265,
+      "learning_rate": 2.111220762292736e-05,
+      "loss": 1.2441,
+      "step": 12444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1995042849301907,
+      "learning_rate": 2.1108087864759114e-05,
+      "loss": 1.4355,
+      "step": 12445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3776040098042723,
+      "learning_rate": 2.1103968214901206e-05,
+      "loss": 1.0166,
+      "step": 12446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.097023390491131,
+      "learning_rate": 2.1099848673468272e-05,
+      "loss": 1.2031,
+      "step": 12447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.44294662402999,
+      "learning_rate": 2.109572924057497e-05,
+      "loss": 1.1035,
+      "step": 12448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2964095078638778,
+      "learning_rate": 2.1091609916335914e-05,
+      "loss": 1.3281,
+      "step": 12449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7111355492093203,
+      "learning_rate": 2.1087490700865778e-05,
+      "loss": 0.9995,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.266416003821975,
+      "learning_rate": 2.108337159427918e-05,
+      "loss": 0.9395,
+      "step": 12451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.2869977321945285,
+      "learning_rate": 2.107925259669076e-05,
+      "loss": 0.8975,
+      "step": 12452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1775228491721945,
+      "learning_rate": 2.1075133708215138e-05,
+      "loss": 1.7148,
+      "step": 12453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6003528475034052,
+      "learning_rate": 2.1071014928966956e-05,
+      "loss": 0.9194,
+      "step": 12454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.361526013522341,
+      "learning_rate": 2.106689625906082e-05,
+      "loss": 0.7415,
+      "step": 12455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.090865514825468,
+      "learning_rate": 2.106277769861137e-05,
+      "loss": 1.1414,
+      "step": 12456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.138864970867051,
+      "learning_rate": 2.1058659247733203e-05,
+      "loss": 0.9395,
+      "step": 12457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.576994355971474,
+      "learning_rate": 2.105454090654094e-05,
+      "loss": 1.0293,
+      "step": 12458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.568788321622915,
+      "learning_rate": 2.105042267514921e-05,
+      "loss": 1.8398,
+      "step": 12459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0456633730936393,
+      "learning_rate": 2.1046304553672606e-05,
+      "loss": 1.5195,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.68058374836996,
+      "learning_rate": 2.104218654222574e-05,
+      "loss": 1.4492,
+      "step": 12461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.500454137986534,
+      "learning_rate": 2.1038068640923207e-05,
+      "loss": 0.9834,
+      "step": 12462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1248475609792408,
+      "learning_rate": 2.1033950849879614e-05,
+      "loss": 1.0977,
+      "step": 12463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.421273926009647,
+      "learning_rate": 2.102983316920955e-05,
+      "loss": 1.2285,
+      "step": 12464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.239265372009427,
+      "learning_rate": 2.1025715599027618e-05,
+      "loss": 1.0542,
+      "step": 12465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5066633116753763,
+      "learning_rate": 2.1021598139448396e-05,
+      "loss": 1.2246,
+      "step": 12466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.54408804407054,
+      "learning_rate": 2.1017480790586487e-05,
+      "loss": 1.1953,
+      "step": 12467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.017691683827722,
+      "learning_rate": 2.1013363552556454e-05,
+      "loss": 1.1445,
+      "step": 12468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.672042518203345,
+      "learning_rate": 2.100924642547289e-05,
+      "loss": 1.1992,
+      "step": 12469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.988723436560625,
+      "learning_rate": 2.100512940945038e-05,
+      "loss": 0.9741,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4206398602941244,
+      "learning_rate": 2.1001012504603497e-05,
+      "loss": 1.1738,
+      "step": 12471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8978002624022947,
+      "learning_rate": 2.09968957110468e-05,
+      "loss": 1.1045,
+      "step": 12472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5858167176404834,
+      "learning_rate": 2.0992779028894875e-05,
+      "loss": 1.4043,
+      "step": 12473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.642411346299236,
+      "learning_rate": 2.098866245826227e-05,
+      "loss": 0.9922,
+      "step": 12474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0192635984162433,
+      "learning_rate": 2.098454599926356e-05,
+      "loss": 0.8206,
+      "step": 12475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.106368647379413,
+      "learning_rate": 2.0980429652013297e-05,
+      "loss": 1.1006,
+      "step": 12476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4376154163129753,
+      "learning_rate": 2.0976313416626037e-05,
+      "loss": 1.2539,
+      "step": 12477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.090158709305049,
+      "learning_rate": 2.0972197293216346e-05,
+      "loss": 0.8975,
+      "step": 12478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.1541694381906975,
+      "learning_rate": 2.0968081281898756e-05,
+      "loss": 1.7739,
+      "step": 12479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.070477274969798,
+      "learning_rate": 2.096396538278783e-05,
+      "loss": 0.6777,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0624325594707384,
+      "learning_rate": 2.09598495959981e-05,
+      "loss": 1.0352,
+      "step": 12481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7884719073718356,
+      "learning_rate": 2.0955733921644116e-05,
+      "loss": 0.7355,
+      "step": 12482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.7225595009562005,
+      "learning_rate": 2.0951618359840403e-05,
+      "loss": 1.2656,
+      "step": 12483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7271851973503694,
+      "learning_rate": 2.094750291070151e-05,
+      "loss": 1.1826,
+      "step": 12484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.275147640016219,
+      "learning_rate": 2.0943387574341957e-05,
+      "loss": 1.1152,
+      "step": 12485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.536390408065478,
+      "learning_rate": 2.0939272350876268e-05,
+      "loss": 1.2285,
+      "step": 12486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.903467392697427,
+      "learning_rate": 2.093515724041899e-05,
+      "loss": 0.7118,
+      "step": 12487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.704819282198777,
+      "learning_rate": 2.0931042243084624e-05,
+      "loss": 1.5469,
+      "step": 12488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.740759061798896,
+      "learning_rate": 2.0926927358987696e-05,
+      "loss": 0.9834,
+      "step": 12489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6583507533284934,
+      "learning_rate": 2.092281258824273e-05,
+      "loss": 1.0527,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.31378241369459,
+      "learning_rate": 2.091869793096422e-05,
+      "loss": 1.5156,
+      "step": 12491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6901370172467427,
+      "learning_rate": 2.0914583387266694e-05,
+      "loss": 1.1445,
+      "step": 12492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.480113182715331,
+      "learning_rate": 2.091046895726464e-05,
+      "loss": 1.0264,
+      "step": 12493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.521205233628973,
+      "learning_rate": 2.0906354641072578e-05,
+      "loss": 0.9121,
+      "step": 12494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2463754493866186,
+      "learning_rate": 2.0902240438804992e-05,
+      "loss": 1.3955,
+      "step": 12495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.558547923089255,
+      "learning_rate": 2.0898126350576386e-05,
+      "loss": 0.8701,
+      "step": 12496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.26804186065499,
+      "learning_rate": 2.089401237650126e-05,
+      "loss": 1.2031,
+      "step": 12497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.959470035501901,
+      "learning_rate": 2.0889898516694095e-05,
+      "loss": 1.6387,
+      "step": 12498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.766900926653327,
+      "learning_rate": 2.0885784771269386e-05,
+      "loss": 0.7451,
+      "step": 12499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.07437206649148,
+      "learning_rate": 2.0881671140341604e-05,
+      "loss": 1.0674,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2860855690801456,
+      "learning_rate": 2.0877557624025246e-05,
+      "loss": 0.8105,
+      "step": 12501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4193885098442047,
+      "learning_rate": 2.0873444222434775e-05,
+      "loss": 1.0957,
+      "step": 12502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8445550806000193,
+      "learning_rate": 2.0869330935684678e-05,
+      "loss": 0.9014,
+      "step": 12503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.367150737615479,
+      "learning_rate": 2.086521776388941e-05,
+      "loss": 0.6702,
+      "step": 12504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.063716529653335,
+      "learning_rate": 2.0861104707163447e-05,
+      "loss": 0.9185,
+      "step": 12505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.00344698288345,
+      "learning_rate": 2.0856991765621266e-05,
+      "loss": 1.1455,
+      "step": 12506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9251687156077866,
+      "learning_rate": 2.085287893937731e-05,
+      "loss": 1.002,
+      "step": 12507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3696748362377256,
+      "learning_rate": 2.084876622854605e-05,
+      "loss": 1.4453,
+      "step": 12508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.243511848385773,
+      "learning_rate": 2.084465363324193e-05,
+      "loss": 0.9497,
+      "step": 12509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.586167139084664,
+      "learning_rate": 2.084054115357941e-05,
+      "loss": 1.2539,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.594008007775434,
+      "learning_rate": 2.083642878967294e-05,
+      "loss": 0.8676,
+      "step": 12511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.030566814786117,
+      "learning_rate": 2.083231654163696e-05,
+      "loss": 1.2065,
+      "step": 12512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.177512228544916,
+      "learning_rate": 2.0828204409585915e-05,
+      "loss": 0.728,
+      "step": 12513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8894407488417808,
+      "learning_rate": 2.0824092393634233e-05,
+      "loss": 0.7139,
+      "step": 12514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8584985670977578,
+      "learning_rate": 2.081998049389637e-05,
+      "loss": 0.73,
+      "step": 12515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.252723952326006,
+      "learning_rate": 2.0815868710486748e-05,
+      "loss": 1.0537,
+      "step": 12516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.57209824111113,
+      "learning_rate": 2.0811757043519792e-05,
+      "loss": 1.1436,
+      "step": 12517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5922484868511226,
+      "learning_rate": 2.0807645493109937e-05,
+      "loss": 0.9814,
+      "step": 12518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.3305828432824,
+      "learning_rate": 2.08035340593716e-05,
+      "loss": 1.2676,
+      "step": 12519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3184092515498897,
+      "learning_rate": 2.07994227424192e-05,
+      "loss": 1.4316,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.488184473380496,
+      "learning_rate": 2.0795311542367154e-05,
+      "loss": 0.5068,
+      "step": 12521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.69068713682149,
+      "learning_rate": 2.0791200459329882e-05,
+      "loss": 0.5469,
+      "step": 12522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.803143257564358,
+      "learning_rate": 2.0787089493421777e-05,
+      "loss": 0.8567,
+      "step": 12523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2645923234140555,
+      "learning_rate": 2.0782978644757262e-05,
+      "loss": 0.5371,
+      "step": 12524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3425173760648024,
+      "learning_rate": 2.077886791345074e-05,
+      "loss": 0.6423,
+      "step": 12525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.276863374641137,
+      "learning_rate": 2.07747572996166e-05,
+      "loss": 1.0283,
+      "step": 12526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.491003251400192,
+      "learning_rate": 2.0770646803369246e-05,
+      "loss": 1.0918,
+      "step": 12527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.446805530999722,
+      "learning_rate": 2.0766536424823066e-05,
+      "loss": 1.2051,
+      "step": 12528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.495337241387485,
+      "learning_rate": 2.0762426164092455e-05,
+      "loss": 1.0996,
+      "step": 12529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8948102922522028,
+      "learning_rate": 2.0758316021291804e-05,
+      "loss": 1.4502,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.726435267586792,
+      "learning_rate": 2.0754205996535488e-05,
+      "loss": 1.3154,
+      "step": 12531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.54906419589234,
+      "learning_rate": 2.0750096089937893e-05,
+      "loss": 0.8369,
+      "step": 12532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.721433082908664,
+      "learning_rate": 2.0745986301613385e-05,
+      "loss": 0.9763,
+      "step": 12533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.432094554807908,
+      "learning_rate": 2.0741876631676348e-05,
+      "loss": 0.8271,
+      "step": 12534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3898769030932274,
+      "learning_rate": 2.0737767080241162e-05,
+      "loss": 0.9473,
+      "step": 12535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9329514795046308,
+      "learning_rate": 2.0733657647422173e-05,
+      "loss": 1.248,
+      "step": 12536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7980259067080446,
+      "learning_rate": 2.0729548333333763e-05,
+      "loss": 0.5398,
+      "step": 12537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7113800127400673,
+      "learning_rate": 2.0725439138090277e-05,
+      "loss": 1.375,
+      "step": 12538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8881188328142,
+      "learning_rate": 2.0721330061806086e-05,
+      "loss": 1.0039,
+      "step": 12539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.610192540705999,
+      "learning_rate": 2.0717221104595534e-05,
+      "loss": 0.3954,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7502807994012937,
+      "learning_rate": 2.0713112266572977e-05,
+      "loss": 0.8564,
+      "step": 12541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6920748075432783,
+      "learning_rate": 2.070900354785275e-05,
+      "loss": 1.1426,
+      "step": 12542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.29974602237546,
+      "learning_rate": 2.0704894948549215e-05,
+      "loss": 1.2822,
+      "step": 12543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2352341287346187,
+      "learning_rate": 2.070078646877671e-05,
+      "loss": 1.2754,
+      "step": 12544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.424148243558583,
+      "learning_rate": 2.069667810864956e-05,
+      "loss": 1.5078,
+      "step": 12545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.196251222521092,
+      "learning_rate": 2.0692569868282115e-05,
+      "loss": 1.0176,
+      "step": 12546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6864914888930183,
+      "learning_rate": 2.068846174778869e-05,
+      "loss": 0.7417,
+      "step": 12547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.2425047640581,
+      "learning_rate": 2.0684353747283626e-05,
+      "loss": 1.4121,
+      "step": 12548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0907590768925632,
+      "learning_rate": 2.0680245866881232e-05,
+      "loss": 0.5788,
+      "step": 12549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.139514607257728,
+      "learning_rate": 2.0676138106695835e-05,
+      "loss": 1.1461,
+      "step": 12550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.134547426487571,
+      "learning_rate": 2.0672030466841755e-05,
+      "loss": 0.7646,
+      "step": 12551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.566797040772844,
+      "learning_rate": 2.0667922947433314e-05,
+      "loss": 1.168,
+      "step": 12552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5275755813471523,
+      "learning_rate": 2.0663815548584806e-05,
+      "loss": 0.6035,
+      "step": 12553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.869844707315015,
+      "learning_rate": 2.0659708270410552e-05,
+      "loss": 0.9961,
+      "step": 12554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2797047200689096,
+      "learning_rate": 2.0655601113024846e-05,
+      "loss": 1.2832,
+      "step": 12555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9291869220810018,
+      "learning_rate": 2.0651494076541997e-05,
+      "loss": 0.8477,
+      "step": 12556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.40389800369549,
+      "learning_rate": 2.0647387161076294e-05,
+      "loss": 1.123,
+      "step": 12557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.352312511597371,
+      "learning_rate": 2.064328036674204e-05,
+      "loss": 0.7654,
+      "step": 12558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.662685362761521,
+      "learning_rate": 2.063917369365351e-05,
+      "loss": 1.4355,
+      "step": 12559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4828553740422827,
+      "learning_rate": 2.0635067141924997e-05,
+      "loss": 1.042,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1779267230427206,
+      "learning_rate": 2.06309607116708e-05,
+      "loss": 0.7363,
+      "step": 12561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9946081663664046,
+      "learning_rate": 2.0626854403005183e-05,
+      "loss": 0.7698,
+      "step": 12562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9961357524674304,
+      "learning_rate": 2.0622748216042434e-05,
+      "loss": 1.6738,
+      "step": 12563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.087869005893643,
+      "learning_rate": 2.0618642150896813e-05,
+      "loss": 0.876,
+      "step": 12564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.41419839479432,
+      "learning_rate": 2.06145362076826e-05,
+      "loss": 1.0488,
+      "step": 12565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.87830148667189,
+      "learning_rate": 2.061043038651406e-05,
+      "loss": 1.1401,
+      "step": 12566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6308282002143297,
+      "learning_rate": 2.0606324687505457e-05,
+      "loss": 1.1426,
+      "step": 12567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.00883842558676,
+      "learning_rate": 2.0602219110771045e-05,
+      "loss": 0.9868,
+      "step": 12568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1678343585175592,
+      "learning_rate": 2.0598113656425083e-05,
+      "loss": 1.2266,
+      "step": 12569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.252095323041391,
+      "learning_rate": 2.0594008324581824e-05,
+      "loss": 0.9561,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.638900464573507,
+      "learning_rate": 2.0589903115355526e-05,
+      "loss": 1.3633,
+      "step": 12571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.322033176707531,
+      "learning_rate": 2.0585798028860425e-05,
+      "loss": 0.8164,
+      "step": 12572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.881632474599601,
+      "learning_rate": 2.0581693065210774e-05,
+      "loss": 0.7012,
+      "step": 12573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.859453449995658,
+      "learning_rate": 2.0577588224520802e-05,
+      "loss": 1.2402,
+      "step": 12574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2283168373077467,
+      "learning_rate": 2.0573483506904752e-05,
+      "loss": 1.052,
+      "step": 12575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9518156832603717,
+      "learning_rate": 2.0569378912476854e-05,
+      "loss": 0.9756,
+      "step": 12576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9430073142728,
+      "learning_rate": 2.056527444135134e-05,
+      "loss": 1.293,
+      "step": 12577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.159668383762625,
+      "learning_rate": 2.0561170093642423e-05,
+      "loss": 0.97,
+      "step": 12578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.619258414286018,
+      "learning_rate": 2.0557065869464336e-05,
+      "loss": 0.4238,
+      "step": 12579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.668930299435965,
+      "learning_rate": 2.0552961768931303e-05,
+      "loss": 1.0186,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.158041902772015,
+      "learning_rate": 2.054885779215753e-05,
+      "loss": 1.7793,
+      "step": 12581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.604503955912367,
+      "learning_rate": 2.054475393925724e-05,
+      "loss": 1.4512,
+      "step": 12582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4672834651902726,
+      "learning_rate": 2.0540650210344625e-05,
+      "loss": 0.9805,
+      "step": 12583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.112070692331466,
+      "learning_rate": 2.0536546605533908e-05,
+      "loss": 0.5415,
+      "step": 12584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 23.182092740478048,
+      "learning_rate": 2.0532443124939275e-05,
+      "loss": 1.6172,
+      "step": 12585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.494090550481856,
+      "learning_rate": 2.0528339768674937e-05,
+      "loss": 1.4883,
+      "step": 12586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5337252423525998,
+      "learning_rate": 2.0524236536855073e-05,
+      "loss": 1.002,
+      "step": 12587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.740587686178113,
+      "learning_rate": 2.0520133429593885e-05,
+      "loss": 0.8848,
+      "step": 12588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1477390773590654,
+      "learning_rate": 2.0516030447005565e-05,
+      "loss": 0.8818,
+      "step": 12589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.474510427334441,
+      "learning_rate": 2.0511927589204293e-05,
+      "loss": 1.1484,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.012811168254506,
+      "learning_rate": 2.0507824856304247e-05,
+      "loss": 0.8828,
+      "step": 12591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.92875171459478,
+      "learning_rate": 2.050372224841961e-05,
+      "loss": 0.9189,
+      "step": 12592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.244503802434392,
+      "learning_rate": 2.0499619765664548e-05,
+      "loss": 1.1211,
+      "step": 12593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.005717244781714,
+      "learning_rate": 2.049551740815324e-05,
+      "loss": 1.2236,
+      "step": 12594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5777021266476394,
+      "learning_rate": 2.049141517599985e-05,
+      "loss": 1.2549,
+      "step": 12595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6505135020682307,
+      "learning_rate": 2.0487313069318535e-05,
+      "loss": 1.3574,
+      "step": 12596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4532448320216766,
+      "learning_rate": 2.048321108822346e-05,
+      "loss": 0.9453,
+      "step": 12597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.7958152171880135,
+      "learning_rate": 2.047910923282878e-05,
+      "loss": 0.9307,
+      "step": 12598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3729883006209773,
+      "learning_rate": 2.0475007503248656e-05,
+      "loss": 1.002,
+      "step": 12599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.182495086904172,
+      "learning_rate": 2.0470905899597225e-05,
+      "loss": 0.8896,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.600094034255106,
+      "learning_rate": 2.0466804421988646e-05,
+      "loss": 0.9978,
+      "step": 12601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.698889263706732,
+      "learning_rate": 2.046270307053705e-05,
+      "loss": 1.0215,
+      "step": 12602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6730425330537493,
+      "learning_rate": 2.0458601845356585e-05,
+      "loss": 0.4877,
+      "step": 12603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.032138806070154,
+      "learning_rate": 2.0454500746561377e-05,
+      "loss": 0.5073,
+      "step": 12604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.695033028644345,
+      "learning_rate": 2.0450399774265567e-05,
+      "loss": 1.6035,
+      "step": 12605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.342978083333413,
+      "learning_rate": 2.0446298928583273e-05,
+      "loss": 0.7402,
+      "step": 12606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.362135146659228,
+      "learning_rate": 2.044219820962862e-05,
+      "loss": 0.8682,
+      "step": 12607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.317526721087512,
+      "learning_rate": 2.0438097617515745e-05,
+      "loss": 1.2197,
+      "step": 12608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.666721156737191,
+      "learning_rate": 2.0433997152358757e-05,
+      "loss": 1.7988,
+      "step": 12609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9226125526965268,
+      "learning_rate": 2.0429896814271767e-05,
+      "loss": 1.0605,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3744112590080717,
+      "learning_rate": 2.0425796603368893e-05,
+      "loss": 0.9375,
+      "step": 12611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3588520000905735,
+      "learning_rate": 2.042169651976423e-05,
+      "loss": 1.0703,
+      "step": 12612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6850697374433468,
+      "learning_rate": 2.0417596563571896e-05,
+      "loss": 0.9268,
+      "step": 12613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.766611666239804,
+      "learning_rate": 2.041349673490598e-05,
+      "loss": 0.5552,
+      "step": 12614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.019908017911391,
+      "learning_rate": 2.0409397033880586e-05,
+      "loss": 0.9717,
+      "step": 12615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0465435141193784,
+      "learning_rate": 2.0405297460609796e-05,
+      "loss": 0.9229,
+      "step": 12616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4391930397970323,
+      "learning_rate": 2.040119801520771e-05,
+      "loss": 1.2568,
+      "step": 12617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4915897047049693,
+      "learning_rate": 2.0397098697788418e-05,
+      "loss": 1.3906,
+      "step": 12618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6319256202832033,
+      "learning_rate": 2.0392999508465987e-05,
+      "loss": 1.4512,
+      "step": 12619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7420378641453844,
+      "learning_rate": 2.0388900447354512e-05,
+      "loss": 0.9233,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7607456404006285,
+      "learning_rate": 2.038480151456805e-05,
+      "loss": 1.2578,
+      "step": 12621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1493966508724123,
+      "learning_rate": 2.0380702710220696e-05,
+      "loss": 0.8613,
+      "step": 12622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4282836821698592,
+      "learning_rate": 2.0376604034426497e-05,
+      "loss": 0.7959,
+      "step": 12623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2121836525086556,
+      "learning_rate": 2.037250548729953e-05,
+      "loss": 0.5269,
+      "step": 12624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.750159219942305,
+      "learning_rate": 2.0368407068953834e-05,
+      "loss": 0.7241,
+      "step": 12625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.1969473799471615,
+      "learning_rate": 2.0364308779503506e-05,
+      "loss": 0.9136,
+      "step": 12626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.198750691720096,
+      "learning_rate": 2.0360210619062564e-05,
+      "loss": 0.8389,
+      "step": 12627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1724798819022033,
+      "learning_rate": 2.035611258774508e-05,
+      "loss": 1.1445,
+      "step": 12628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3129417556722984,
+      "learning_rate": 2.035201468566509e-05,
+      "loss": 1.0234,
+      "step": 12629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.910049299822008,
+      "learning_rate": 2.034791691293664e-05,
+      "loss": 1.394,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5633016123336394,
+      "learning_rate": 2.0343819269673764e-05,
+      "loss": 0.9473,
+      "step": 12631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.429754225621149,
+      "learning_rate": 2.033972175599051e-05,
+      "loss": 1.4961,
+      "step": 12632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0163149576903794,
+      "learning_rate": 2.0335624372000893e-05,
+      "loss": 0.4182,
+      "step": 12633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.190898825441613,
+      "learning_rate": 2.033152711781896e-05,
+      "loss": 1.1855,
+      "step": 12634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7885237636421811,
+      "learning_rate": 2.0327429993558718e-05,
+      "loss": 1.1147,
+      "step": 12635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.292690568534302,
+      "learning_rate": 2.0323332999334198e-05,
+      "loss": 1.6738,
+      "step": 12636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0693137216325344,
+      "learning_rate": 2.0319236135259424e-05,
+      "loss": 0.8916,
+      "step": 12637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8225602679242217,
+      "learning_rate": 2.0315139401448394e-05,
+      "loss": 0.4707,
+      "step": 12638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.208100888476658,
+      "learning_rate": 2.0311042798015138e-05,
+      "loss": 1.3779,
+      "step": 12639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1800805834653985,
+      "learning_rate": 2.0306946325073643e-05,
+      "loss": 1.0156,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.646215716587251,
+      "learning_rate": 2.0302849982737928e-05,
+      "loss": 1.2812,
+      "step": 12641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.258880189408922,
+      "learning_rate": 2.0298753771121977e-05,
+      "loss": 0.9297,
+      "step": 12642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.195588444102228,
+      "learning_rate": 2.0294657690339804e-05,
+      "loss": 0.4912,
+      "step": 12643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9988259120429595,
+      "learning_rate": 2.0290561740505383e-05,
+      "loss": 1.2207,
+      "step": 12644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4003505967281393,
+      "learning_rate": 2.0286465921732712e-05,
+      "loss": 1.2734,
+      "step": 12645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.952625212182495,
+      "learning_rate": 2.028237023413578e-05,
+      "loss": 0.3684,
+      "step": 12646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7239151046172245,
+      "learning_rate": 2.0278274677828562e-05,
+      "loss": 0.7119,
+      "step": 12647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7535818355293666,
+      "learning_rate": 2.027417925292504e-05,
+      "loss": 1.0381,
+      "step": 12648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1722601295013044,
+      "learning_rate": 2.0270083959539188e-05,
+      "loss": 0.8857,
+      "step": 12649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.993884369894785,
+      "learning_rate": 2.0265988797784967e-05,
+      "loss": 0.5739,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.242402503895336,
+      "learning_rate": 2.0261893767776356e-05,
+      "loss": 1.918,
+      "step": 12651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.277907640144007,
+      "learning_rate": 2.025779886962731e-05,
+      "loss": 0.8262,
+      "step": 12652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1685004864030595,
+      "learning_rate": 2.0253704103451783e-05,
+      "loss": 1.1377,
+      "step": 12653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.507581809429583,
+      "learning_rate": 2.024960946936375e-05,
+      "loss": 0.8174,
+      "step": 12654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0271005355978025,
+      "learning_rate": 2.0245514967477146e-05,
+      "loss": 1.0488,
+      "step": 12655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.866043637846792,
+      "learning_rate": 2.024142059790593e-05,
+      "loss": 1.6641,
+      "step": 12656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8977450139976857,
+      "learning_rate": 2.023732636076404e-05,
+      "loss": 0.4109,
+      "step": 12657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2601722081158675,
+      "learning_rate": 2.0233232256165422e-05,
+      "loss": 0.8818,
+      "step": 12658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8899789296734273,
+      "learning_rate": 2.0229138284224002e-05,
+      "loss": 1.3945,
+      "step": 12659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.991924573558553,
+      "learning_rate": 2.0225044445053727e-05,
+      "loss": 1.1133,
+      "step": 12660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.503326967928024,
+      "learning_rate": 2.0220950738768518e-05,
+      "loss": 0.5037,
+      "step": 12661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7791122621837916,
+      "learning_rate": 2.0216857165482298e-05,
+      "loss": 0.9678,
+      "step": 12662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.268395959415256,
+      "learning_rate": 2.0212763725309006e-05,
+      "loss": 0.9097,
+      "step": 12663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7031002217861397,
+      "learning_rate": 2.0208670418362546e-05,
+      "loss": 0.5547,
+      "step": 12664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.1702514232337835,
+      "learning_rate": 2.0204577244756838e-05,
+      "loss": 0.6514,
+      "step": 12665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1454933662605367,
+      "learning_rate": 2.0200484204605796e-05,
+      "loss": 1.0144,
+      "step": 12666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7525367307507596,
+      "learning_rate": 2.0196391298023322e-05,
+      "loss": 0.9648,
+      "step": 12667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.545882517271378,
+      "learning_rate": 2.0192298525123325e-05,
+      "loss": 1.0908,
+      "step": 12668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.80652084572827,
+      "learning_rate": 2.01882058860197e-05,
+      "loss": 0.9102,
+      "step": 12669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.314718809016285,
+      "learning_rate": 2.0184113380826353e-05,
+      "loss": 1.3379,
+      "step": 12670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3124530890578123,
+      "learning_rate": 2.0180021009657162e-05,
+      "loss": 0.8398,
+      "step": 12671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.993902275329934,
+      "learning_rate": 2.017592877262602e-05,
+      "loss": 1.2188,
+      "step": 12672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3041214587264536,
+      "learning_rate": 2.017183666984683e-05,
+      "loss": 0.4006,
+      "step": 12673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.527509971568972,
+      "learning_rate": 2.016774470143345e-05,
+      "loss": 1.2852,
+      "step": 12674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.482874477554733,
+      "learning_rate": 2.016365286749977e-05,
+      "loss": 0.8911,
+      "step": 12675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.005036521106941,
+      "learning_rate": 2.0159561168159665e-05,
+      "loss": 1.0151,
+      "step": 12676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.962244664578706,
+      "learning_rate": 2.0155469603527003e-05,
+      "loss": 1.5742,
+      "step": 12677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.493314482248751,
+      "learning_rate": 2.0151378173715647e-05,
+      "loss": 1.0234,
+      "step": 12678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.1783184941603,
+      "learning_rate": 2.0147286878839462e-05,
+      "loss": 1.3477,
+      "step": 12679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0318101550725904,
+      "learning_rate": 2.014319571901231e-05,
+      "loss": 1.0039,
+      "step": 12680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7771305050650907,
+      "learning_rate": 2.013910469434803e-05,
+      "loss": 1.4395,
+      "step": 12681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0062200913557375,
+      "learning_rate": 2.0135013804960504e-05,
+      "loss": 1.123,
+      "step": 12682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4093314912778774,
+      "learning_rate": 2.0130923050963557e-05,
+      "loss": 1.4785,
+      "step": 12683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.570395610358452,
+      "learning_rate": 2.0126832432471044e-05,
+      "loss": 1.1152,
+      "step": 12684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8713825521733067,
+      "learning_rate": 2.0122741949596797e-05,
+      "loss": 1.6582,
+      "step": 12685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.296773526341965,
+      "learning_rate": 2.0118651602454654e-05,
+      "loss": 0.7607,
+      "step": 12686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8377434311018535,
+      "learning_rate": 2.0114561391158455e-05,
+      "loss": 0.7383,
+      "step": 12687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6487047628848623,
+      "learning_rate": 2.011047131582202e-05,
+      "loss": 1.0557,
+      "step": 12688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.172766962011738,
+      "learning_rate": 2.010638137655918e-05,
+      "loss": 0.7527,
+      "step": 12689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7055882162341693,
+      "learning_rate": 2.0102291573483746e-05,
+      "loss": 0.771,
+      "step": 12690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.780961206930776,
+      "learning_rate": 2.009820190670955e-05,
+      "loss": 1.0166,
+      "step": 12691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.818007438611947,
+      "learning_rate": 2.0094112376350404e-05,
+      "loss": 1.2715,
+      "step": 12692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.261593314859792,
+      "learning_rate": 2.0090022982520105e-05,
+      "loss": 0.5469,
+      "step": 12693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8256131814845773,
+      "learning_rate": 2.0085933725332474e-05,
+      "loss": 0.5942,
+      "step": 12694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3506487214753466,
+      "learning_rate": 2.0081844604901303e-05,
+      "loss": 0.9319,
+      "step": 12695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7952111265800625,
+      "learning_rate": 2.0077755621340398e-05,
+      "loss": 0.9375,
+      "step": 12696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.64297462077085,
+      "learning_rate": 2.0073666774763548e-05,
+      "loss": 1.1523,
+      "step": 12697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.761348543079484,
+      "learning_rate": 2.006957806528455e-05,
+      "loss": 1.0342,
+      "step": 12698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.26872676147,
+      "learning_rate": 2.0065489493017176e-05,
+      "loss": 1.6777,
+      "step": 12699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5974544131231974,
+      "learning_rate": 2.006140105807523e-05,
+      "loss": 1.0273,
+      "step": 12700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.665328522303451,
+      "learning_rate": 2.005731276057248e-05,
+      "loss": 1.2305,
+      "step": 12701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0941638091800425,
+      "learning_rate": 2.0053224600622703e-05,
+      "loss": 1.2,
+      "step": 12702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8987773700923778,
+      "learning_rate": 2.0049136578339676e-05,
+      "loss": 0.792,
+      "step": 12703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6440442739265684,
+      "learning_rate": 2.0045048693837156e-05,
+      "loss": 0.874,
+      "step": 12704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.549672374996073,
+      "learning_rate": 2.004096094722892e-05,
+      "loss": 0.6113,
+      "step": 12705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.936608199433486,
+      "learning_rate": 2.0036873338628716e-05,
+      "loss": 1.0703,
+      "step": 12706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.654668090102719,
+      "learning_rate": 2.0032785868150307e-05,
+      "loss": 1.207,
+      "step": 12707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8272685234642947,
+      "learning_rate": 2.0028698535907454e-05,
+      "loss": 0.5894,
+      "step": 12708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.041279668008677,
+      "learning_rate": 2.002461134201388e-05,
+      "loss": 1.3223,
+      "step": 12709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3628391591158517,
+      "learning_rate": 2.0020524286583357e-05,
+      "loss": 0.6642,
+      "step": 12710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.570105825538663,
+      "learning_rate": 2.001643736972962e-05,
+      "loss": 1.2109,
+      "step": 12711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.634104902837041,
+      "learning_rate": 2.0012350591566396e-05,
+      "loss": 1.3691,
+      "step": 12712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.013293670154137,
+      "learning_rate": 2.0008263952207427e-05,
+      "loss": 1.1562,
+      "step": 12713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.685062639434483,
+      "learning_rate": 2.000417745176644e-05,
+      "loss": 0.8662,
+      "step": 12714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.580582863870328,
+      "learning_rate": 2.0000091090357166e-05,
+      "loss": 1.6914,
+      "step": 12715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5779977301072776,
+      "learning_rate": 1.9996004868093314e-05,
+      "loss": 1.0488,
+      "step": 12716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.9403765414721565,
+      "learning_rate": 1.9991918785088615e-05,
+      "loss": 1.2734,
+      "step": 12717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.43415706601699,
+      "learning_rate": 1.9987832841456766e-05,
+      "loss": 1.2471,
+      "step": 12718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.212686554538454,
+      "learning_rate": 1.998374703731149e-05,
+      "loss": 1.252,
+      "step": 12719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.689055121268874,
+      "learning_rate": 1.9979661372766506e-05,
+      "loss": 0.8613,
+      "step": 12720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.535288567756273,
+      "learning_rate": 1.9975575847935492e-05,
+      "loss": 1.3682,
+      "step": 12721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.505551431461682,
+      "learning_rate": 1.9971490462932162e-05,
+      "loss": 1.626,
+      "step": 12722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.526569418818072,
+      "learning_rate": 1.99674052178702e-05,
+      "loss": 1.291,
+      "step": 12723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8364857827595187,
+      "learning_rate": 1.996332011286331e-05,
+      "loss": 0.6914,
+      "step": 12724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.788859210647065,
+      "learning_rate": 1.9959235148025163e-05,
+      "loss": 1.0649,
+      "step": 12725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.65344269778674,
+      "learning_rate": 1.9955150323469453e-05,
+      "loss": 1.3652,
+      "step": 12726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.89921752633614,
+      "learning_rate": 1.9951065639309844e-05,
+      "loss": 1.2051,
+      "step": 12727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7597335737976496,
+      "learning_rate": 1.994698109566004e-05,
+      "loss": 1.0215,
+      "step": 12728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.039774810362082,
+      "learning_rate": 1.9942896692633685e-05,
+      "loss": 1.0059,
+      "step": 12729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.166362671471471,
+      "learning_rate": 1.9938812430344463e-05,
+      "loss": 0.606,
+      "step": 12730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9884415057937965,
+      "learning_rate": 1.9934728308906025e-05,
+      "loss": 1.2451,
+      "step": 12731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.39313669340834,
+      "learning_rate": 1.993064432843204e-05,
+      "loss": 0.6641,
+      "step": 12732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.26867544352042,
+      "learning_rate": 1.992656048903616e-05,
+      "loss": 0.9785,
+      "step": 12733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6218513648749395,
+      "learning_rate": 1.9922476790832036e-05,
+      "loss": 0.8673,
+      "step": 12734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.062954014600767,
+      "learning_rate": 1.9918393233933312e-05,
+      "loss": 0.9661,
+      "step": 12735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5864446277266047,
+      "learning_rate": 1.9914309818453633e-05,
+      "loss": 1.2676,
+      "step": 12736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8014406002520795,
+      "learning_rate": 1.991022654450665e-05,
+      "loss": 1.0767,
+      "step": 12737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.79639305651018,
+      "learning_rate": 1.9906143412205985e-05,
+      "loss": 1.0762,
+      "step": 12738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5600769868837876,
+      "learning_rate": 1.990206042166528e-05,
+      "loss": 0.7168,
+      "step": 12739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.029036276337065,
+      "learning_rate": 1.9897977572998153e-05,
+      "loss": 1.5078,
+      "step": 12740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.512615579176227,
+      "learning_rate": 1.9893894866318237e-05,
+      "loss": 1.3633,
+      "step": 12741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7702225224599,
+      "learning_rate": 1.9889812301739143e-05,
+      "loss": 1.1846,
+      "step": 12742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9721041973898208,
+      "learning_rate": 1.9885729879374503e-05,
+      "loss": 1.1455,
+      "step": 12743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.291971174874387,
+      "learning_rate": 1.9881647599337906e-05,
+      "loss": 0.4919,
+      "step": 12744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.262323433746416,
+      "learning_rate": 1.987756546174298e-05,
+      "loss": 1.0601,
+      "step": 12745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8226679023667793,
+      "learning_rate": 1.9873483466703308e-05,
+      "loss": 0.9092,
+      "step": 12746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.068422518024128,
+      "learning_rate": 1.9869401614332518e-05,
+      "loss": 0.9019,
+      "step": 12747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.668999507809406,
+      "learning_rate": 1.9865319904744188e-05,
+      "loss": 0.5786,
+      "step": 12748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6344517887820595,
+      "learning_rate": 1.9861238338051923e-05,
+      "loss": 1.0986,
+      "step": 12749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0689661699595123,
+      "learning_rate": 1.985715691436929e-05,
+      "loss": 1.3828,
+      "step": 12750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.917066762561329,
+      "learning_rate": 1.98530756338099e-05,
+      "loss": 0.7612,
+      "step": 12751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2909063756903767,
+      "learning_rate": 1.984899449648731e-05,
+      "loss": 1.5879,
+      "step": 12752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9038279295245975,
+      "learning_rate": 1.9844913502515112e-05,
+      "loss": 1.0742,
+      "step": 12753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.990700083760775,
+      "learning_rate": 1.984083265200687e-05,
+      "loss": 1.0938,
+      "step": 12754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.444015890422552,
+      "learning_rate": 1.9836751945076142e-05,
+      "loss": 0.6758,
+      "step": 12755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.926563758106083,
+      "learning_rate": 1.9832671381836525e-05,
+      "loss": 0.5786,
+      "step": 12756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.481784933930108,
+      "learning_rate": 1.982859096240155e-05,
+      "loss": 0.8489,
+      "step": 12757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9448147601403654,
+      "learning_rate": 1.982451068688479e-05,
+      "loss": 1.3555,
+      "step": 12758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8592564292462535,
+      "learning_rate": 1.9820430555399784e-05,
+      "loss": 0.9785,
+      "step": 12759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.352048960845356,
+      "learning_rate": 1.9816350568060094e-05,
+      "loss": 1.3809,
+      "step": 12760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5465077328477466,
+      "learning_rate": 1.981227072497925e-05,
+      "loss": 0.5579,
+      "step": 12761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.396415131181789,
+      "learning_rate": 1.9808191026270805e-05,
+      "loss": 0.7236,
+      "step": 12762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9371844619257008,
+      "learning_rate": 1.9804111472048286e-05,
+      "loss": 1.3052,
+      "step": 12763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.641807199791061,
+      "learning_rate": 1.9800032062425228e-05,
+      "loss": 1.002,
+      "step": 12764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8198874308807895,
+      "learning_rate": 1.979595279751516e-05,
+      "loss": 1.1592,
+      "step": 12765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4997389559808267,
+      "learning_rate": 1.9791873677431617e-05,
+      "loss": 1.5176,
+      "step": 12766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.68796987639873,
+      "learning_rate": 1.9787794702288102e-05,
+      "loss": 1.5762,
+      "step": 12767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4778376524533003,
+      "learning_rate": 1.9783715872198143e-05,
+      "loss": 1.0166,
+      "step": 12768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7561685292091471,
+      "learning_rate": 1.9779637187275245e-05,
+      "loss": 0.7793,
+      "step": 12769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.488538791617043,
+      "learning_rate": 1.977555864763292e-05,
+      "loss": 0.9619,
+      "step": 12770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.340690141288337,
+      "learning_rate": 1.9771480253384665e-05,
+      "loss": 1.3711,
+      "step": 12771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2305122935302233,
+      "learning_rate": 1.9767402004644e-05,
+      "loss": 1.5566,
+      "step": 12772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.236785015759466,
+      "learning_rate": 1.9763323901524393e-05,
+      "loss": 0.5098,
+      "step": 12773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.960592043409149,
+      "learning_rate": 1.9759245944139347e-05,
+      "loss": 0.9648,
+      "step": 12774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6942826667101514,
+      "learning_rate": 1.9755168132602363e-05,
+      "loss": 0.9937,
+      "step": 12775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9298997218803997,
+      "learning_rate": 1.975109046702691e-05,
+      "loss": 0.9937,
+      "step": 12776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1381818843473055,
+      "learning_rate": 1.974701294752648e-05,
+      "loss": 0.915,
+      "step": 12777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9395501149983436,
+      "learning_rate": 1.9742935574214534e-05,
+      "loss": 0.9326,
+      "step": 12778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.228534454346395,
+      "learning_rate": 1.9738858347204558e-05,
+      "loss": 1.2832,
+      "step": 12779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.706379217676384,
+      "learning_rate": 1.9734781266610004e-05,
+      "loss": 0.8359,
+      "step": 12780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.102214907965771,
+      "learning_rate": 1.9730704332544354e-05,
+      "loss": 1.0054,
+      "step": 12781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2106093308031243,
+      "learning_rate": 1.9726627545121046e-05,
+      "loss": 0.8057,
+      "step": 12782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.38396768553166,
+      "learning_rate": 1.9722550904453544e-05,
+      "loss": 0.6816,
+      "step": 12783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.143130652819877,
+      "learning_rate": 1.9718474410655308e-05,
+      "loss": 0.9502,
+      "step": 12784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1954177055667254,
+      "learning_rate": 1.9714398063839786e-05,
+      "loss": 1.1035,
+      "step": 12785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.563662195659676,
+      "learning_rate": 1.97103218641204e-05,
+      "loss": 1.1445,
+      "step": 12786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5081006416375646,
+      "learning_rate": 1.970624581161062e-05,
+      "loss": 1.3262,
+      "step": 12787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.801347169972866,
+      "learning_rate": 1.970216990642385e-05,
+      "loss": 1.3965,
+      "step": 12788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0390836634213625,
+      "learning_rate": 1.9698094148673544e-05,
+      "loss": 0.5696,
+      "step": 12789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.045030563861445,
+      "learning_rate": 1.9694018538473113e-05,
+      "loss": 1.0527,
+      "step": 12790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.179502679178908,
+      "learning_rate": 1.9689943075935992e-05,
+      "loss": 0.655,
+      "step": 12791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.501686507296983,
+      "learning_rate": 1.9685867761175584e-05,
+      "loss": 0.9111,
+      "step": 12792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.535238433636975,
+      "learning_rate": 1.9681792594305316e-05,
+      "loss": 1.0029,
+      "step": 12793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.541852197433475,
+      "learning_rate": 1.9677717575438602e-05,
+      "loss": 0.9897,
+      "step": 12794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0106692641356982,
+      "learning_rate": 1.9673642704688832e-05,
+      "loss": 1.2031,
+      "step": 12795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.956523415853687,
+      "learning_rate": 1.966956798216943e-05,
+      "loss": 1.5566,
+      "step": 12796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6579950322663763,
+      "learning_rate": 1.966549340799377e-05,
+      "loss": 1.123,
+      "step": 12797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.419753591429777,
+      "learning_rate": 1.9661418982275263e-05,
+      "loss": 0.8032,
+      "step": 12798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1671420333922526,
+      "learning_rate": 1.9657344705127288e-05,
+      "loss": 0.8818,
+      "step": 12799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1110848737502823,
+      "learning_rate": 1.9653270576663237e-05,
+      "loss": 0.7188,
+      "step": 12800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.0648707737305445,
+      "learning_rate": 1.9649196596996482e-05,
+      "loss": 0.9424,
+      "step": 12801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3623099356597344,
+      "learning_rate": 1.9645122766240413e-05,
+      "loss": 1.0562,
+      "step": 12802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6182178803390297,
+      "learning_rate": 1.9641049084508394e-05,
+      "loss": 1.0044,
+      "step": 12803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.238785100560441,
+      "learning_rate": 1.9636975551913804e-05,
+      "loss": 0.451,
+      "step": 12804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3861512853427733,
+      "learning_rate": 1.963290216857e-05,
+      "loss": 1.1484,
+      "step": 12805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.263436148007263,
+      "learning_rate": 1.9628828934590347e-05,
+      "loss": 0.7666,
+      "step": 12806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6432280178036445,
+      "learning_rate": 1.9624755850088187e-05,
+      "loss": 1.123,
+      "step": 12807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.710326584171002,
+      "learning_rate": 1.9620682915176898e-05,
+      "loss": 1.582,
+      "step": 12808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.679201908021993,
+      "learning_rate": 1.9616610129969805e-05,
+      "loss": 0.5078,
+      "step": 12809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0830956616601686,
+      "learning_rate": 1.961253749458026e-05,
+      "loss": 0.709,
+      "step": 12810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9209957437542575,
+      "learning_rate": 1.9608465009121597e-05,
+      "loss": 0.717,
+      "step": 12811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.347614196949036,
+      "learning_rate": 1.9604392673707163e-05,
+      "loss": 1.6738,
+      "step": 12812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7645781395430733,
+      "learning_rate": 1.960032048845029e-05,
+      "loss": 0.6855,
+      "step": 12813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.315976493814988,
+      "learning_rate": 1.9596248453464294e-05,
+      "loss": 1.0137,
+      "step": 12814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4017364975013757,
+      "learning_rate": 1.9592176568862512e-05,
+      "loss": 1.3389,
+      "step": 12815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.425021184034339,
+      "learning_rate": 1.9588104834758244e-05,
+      "loss": 0.8789,
+      "step": 12816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8655806468741267,
+      "learning_rate": 1.9584033251264824e-05,
+      "loss": 1.2754,
+      "step": 12817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.060939078130705,
+      "learning_rate": 1.957996181849555e-05,
+      "loss": 0.8613,
+      "step": 12818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7133906879024967,
+      "learning_rate": 1.9575890536563733e-05,
+      "loss": 1.3833,
+      "step": 12819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7510192757428786,
+      "learning_rate": 1.9571819405582665e-05,
+      "loss": 0.8135,
+      "step": 12820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.997640629168108,
+      "learning_rate": 1.956774842566566e-05,
+      "loss": 1.0156,
+      "step": 12821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7326274410031166,
+      "learning_rate": 1.9563677596926003e-05,
+      "loss": 0.835,
+      "step": 12822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.367886581681753,
+      "learning_rate": 1.9559606919476993e-05,
+      "loss": 0.7207,
+      "step": 12823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.463445921222238,
+      "learning_rate": 1.95555363934319e-05,
+      "loss": 0.8684,
+      "step": 12824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7241664731583906,
+      "learning_rate": 1.9551466018904018e-05,
+      "loss": 1.1621,
+      "step": 12825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.181243549792414,
+      "learning_rate": 1.9547395796006613e-05,
+      "loss": 1.1699,
+      "step": 12826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8507569568693423,
+      "learning_rate": 1.954332572485297e-05,
+      "loss": 1.0242,
+      "step": 12827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6848837959816443,
+      "learning_rate": 1.9539255805556344e-05,
+      "loss": 0.5745,
+      "step": 12828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5529604799265204,
+      "learning_rate": 1.9535186038230003e-05,
+      "loss": 1.1777,
+      "step": 12829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.311745467711508,
+      "learning_rate": 1.953111642298722e-05,
+      "loss": 1.3281,
+      "step": 12830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6058110551295544,
+      "learning_rate": 1.9527046959941233e-05,
+      "loss": 0.918,
+      "step": 12831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.44602190641386,
+      "learning_rate": 1.952297764920531e-05,
+      "loss": 0.7852,
+      "step": 12832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.073894123314109,
+      "learning_rate": 1.951890849089268e-05,
+      "loss": 0.473,
+      "step": 12833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9903777635187774,
+      "learning_rate": 1.95148394851166e-05,
+      "loss": 1.0234,
+      "step": 12834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.109493709200107,
+      "learning_rate": 1.9510770631990304e-05,
+      "loss": 0.8066,
+      "step": 12835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5570723574906657,
+      "learning_rate": 1.950670193162703e-05,
+      "loss": 1.2344,
+      "step": 12836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3919356997851633,
+      "learning_rate": 1.9502633384139994e-05,
+      "loss": 1.291,
+      "step": 12837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8070007815195157,
+      "learning_rate": 1.9498564989642435e-05,
+      "loss": 1.0698,
+      "step": 12838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2601486844128367,
+      "learning_rate": 1.949449674824758e-05,
+      "loss": 0.7246,
+      "step": 12839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.580482305376491,
+      "learning_rate": 1.9490428660068634e-05,
+      "loss": 0.9688,
+      "step": 12840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1840901637975967,
+      "learning_rate": 1.9486360725218823e-05,
+      "loss": 0.876,
+      "step": 12841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.965828545096833,
+      "learning_rate": 1.9482292943811343e-05,
+      "loss": 1.2705,
+      "step": 12842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.544297300504981,
+      "learning_rate": 1.94782253159594e-05,
+      "loss": 1.3027,
+      "step": 12843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4783892343510687,
+      "learning_rate": 1.947415784177621e-05,
+      "loss": 0.7646,
+      "step": 12844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2018306528260445,
+      "learning_rate": 1.9470090521374947e-05,
+      "loss": 0.4353,
+      "step": 12845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.871627253725762,
+      "learning_rate": 1.9466023354868816e-05,
+      "loss": 0.6748,
+      "step": 12846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2948930973927073,
+      "learning_rate": 1.9461956342371006e-05,
+      "loss": 1.8105,
+      "step": 12847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7870331294570276,
+      "learning_rate": 1.9457889483994686e-05,
+      "loss": 0.9814,
+      "step": 12848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.122925250404482,
+      "learning_rate": 1.9453822779853057e-05,
+      "loss": 1.0479,
+      "step": 12849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.530339677798276,
+      "learning_rate": 1.9449756230059272e-05,
+      "loss": 0.9058,
+      "step": 12850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8633374859462772,
+      "learning_rate": 1.9445689834726523e-05,
+      "loss": 1.1758,
+      "step": 12851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2411947677793607,
+      "learning_rate": 1.944162359396796e-05,
+      "loss": 1.2617,
+      "step": 12852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.989400372572592,
+      "learning_rate": 1.9437557507896755e-05,
+      "loss": 0.8701,
+      "step": 12853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7228825172748197,
+      "learning_rate": 1.943349157662605e-05,
+      "loss": 1.5508,
+      "step": 12854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.275070914536119,
+      "learning_rate": 1.9429425800269016e-05,
+      "loss": 1.0127,
+      "step": 12855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.337230812536476,
+      "learning_rate": 1.942536017893879e-05,
+      "loss": 1.2559,
+      "step": 12856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.978855873206947,
+      "learning_rate": 1.9421294712748515e-05,
+      "loss": 0.9595,
+      "step": 12857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.715813740061945,
+      "learning_rate": 1.9417229401811346e-05,
+      "loss": 1.0898,
+      "step": 12858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.670517383876327,
+      "learning_rate": 1.9413164246240407e-05,
+      "loss": 2.1055,
+      "step": 12859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3875320320444855,
+      "learning_rate": 1.9409099246148836e-05,
+      "loss": 0.979,
+      "step": 12860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8741019654254307,
+      "learning_rate": 1.9405034401649753e-05,
+      "loss": 1.0039,
+      "step": 12861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.787044828206214,
+      "learning_rate": 1.9400969712856282e-05,
+      "loss": 1.2188,
+      "step": 12862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.509005541219002,
+      "learning_rate": 1.9396905179881552e-05,
+      "loss": 0.9009,
+      "step": 12863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.115409336797424,
+      "learning_rate": 1.9392840802838666e-05,
+      "loss": 1.0156,
+      "step": 12864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1988108715763626,
+      "learning_rate": 1.9388776581840742e-05,
+      "loss": 1.1914,
+      "step": 12865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.533140912611889,
+      "learning_rate": 1.9384712517000868e-05,
+      "loss": 0.9717,
+      "step": 12866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9915957899796517,
+      "learning_rate": 1.938064860843217e-05,
+      "loss": 1.5098,
+      "step": 12867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9447691131684335,
+      "learning_rate": 1.9376584856247734e-05,
+      "loss": 1.29,
+      "step": 12868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5277677938031338,
+      "learning_rate": 1.937252126056065e-05,
+      "loss": 0.7861,
+      "step": 12869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.267656115205555,
+      "learning_rate": 1.936845782148401e-05,
+      "loss": 1.8496,
+      "step": 12870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.182756785594445,
+      "learning_rate": 1.936439453913089e-05,
+      "loss": 0.8394,
+      "step": 12871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.523963330167221,
+      "learning_rate": 1.9360331413614384e-05,
+      "loss": 1.2207,
+      "step": 12872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9486969457699796,
+      "learning_rate": 1.9356268445047555e-05,
+      "loss": 1.2607,
+      "step": 12873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6557051284195445,
+      "learning_rate": 1.935220563354348e-05,
+      "loss": 0.5957,
+      "step": 12874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0564854006612836,
+      "learning_rate": 1.9348142979215217e-05,
+      "loss": 1.7559,
+      "step": 12875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7000124436550905,
+      "learning_rate": 1.9344080482175835e-05,
+      "loss": 1.1162,
+      "step": 12876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.039323396207031,
+      "learning_rate": 1.9340018142538396e-05,
+      "loss": 1.4338,
+      "step": 12877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.862479646236656,
+      "learning_rate": 1.9335955960415947e-05,
+      "loss": 1.8457,
+      "step": 12878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0732386188710614,
+      "learning_rate": 1.9331893935921538e-05,
+      "loss": 0.8477,
+      "step": 12879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.280017981803411,
+      "learning_rate": 1.9327832069168208e-05,
+      "loss": 0.9648,
+      "step": 12880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.091658273909541,
+      "learning_rate": 1.9323770360269012e-05,
+      "loss": 1.0791,
+      "step": 12881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.232945871596532,
+      "learning_rate": 1.9319708809336968e-05,
+      "loss": 1.5625,
+      "step": 12882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.431224630307423,
+      "learning_rate": 1.9315647416485115e-05,
+      "loss": 1.2588,
+      "step": 12883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.7885178529850085,
+      "learning_rate": 1.9311586181826486e-05,
+      "loss": 1.3184,
+      "step": 12884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1905683820672097,
+      "learning_rate": 1.9307525105474087e-05,
+      "loss": 0.5977,
+      "step": 12885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8141006153666543,
+      "learning_rate": 1.9303464187540952e-05,
+      "loss": 0.9902,
+      "step": 12886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.65226475642985,
+      "learning_rate": 1.9299403428140096e-05,
+      "loss": 1.4434,
+      "step": 12887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.1297655976340595,
+      "learning_rate": 1.9295342827384517e-05,
+      "loss": 1.0303,
+      "step": 12888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5290073261222776,
+      "learning_rate": 1.9291282385387228e-05,
+      "loss": 1.3652,
+      "step": 12889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5421140201251826,
+      "learning_rate": 1.928722210226122e-05,
+      "loss": 0.6096,
+      "step": 12890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.305896734577221,
+      "learning_rate": 1.9283161978119504e-05,
+      "loss": 0.9873,
+      "step": 12891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.888866487644429,
+      "learning_rate": 1.9279102013075057e-05,
+      "loss": 0.7109,
+      "step": 12892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6197488008410217,
+      "learning_rate": 1.9275042207240877e-05,
+      "loss": 0.7002,
+      "step": 12893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.940736975650991,
+      "learning_rate": 1.927098256072993e-05,
+      "loss": 1.3213,
+      "step": 12894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.4013663929725055,
+      "learning_rate": 1.926692307365521e-05,
+      "loss": 1.1348,
+      "step": 12895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.118985810862492,
+      "learning_rate": 1.9262863746129694e-05,
+      "loss": 1.625,
+      "step": 12896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.551271540444783,
+      "learning_rate": 1.925880457826634e-05,
+      "loss": 0.5889,
+      "step": 12897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.447954682815738,
+      "learning_rate": 1.9254745570178125e-05,
+      "loss": 0.6865,
+      "step": 12898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.05246281093169,
+      "learning_rate": 1.925068672197799e-05,
+      "loss": 1.0664,
+      "step": 12899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3648302282646854,
+      "learning_rate": 1.924662803377891e-05,
+      "loss": 1.373,
+      "step": 12900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.695188923089853,
+      "learning_rate": 1.9242569505693828e-05,
+      "loss": 1.5645,
+      "step": 12901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7250964721885818,
+      "learning_rate": 1.9238511137835692e-05,
+      "loss": 0.8535,
+      "step": 12902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.297324000477814,
+      "learning_rate": 1.9234452930317437e-05,
+      "loss": 1.2754,
+      "step": 12903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6541572854092346,
+      "learning_rate": 1.9230394883252024e-05,
+      "loss": 0.5259,
+      "step": 12904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.527223468594711,
+      "learning_rate": 1.9226336996752364e-05,
+      "loss": 1.3242,
+      "step": 12905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.834872269854549,
+      "learning_rate": 1.9222279270931403e-05,
+      "loss": 1.1006,
+      "step": 12906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.174730559466158,
+      "learning_rate": 1.921822170590205e-05,
+      "loss": 1.2754,
+      "step": 12907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2446065312081216,
+      "learning_rate": 1.921416430177724e-05,
+      "loss": 1.1006,
+      "step": 12908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.906401893942785,
+      "learning_rate": 1.9210107058669876e-05,
+      "loss": 1.0029,
+      "step": 12909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.609856221569705,
+      "learning_rate": 1.9206049976692884e-05,
+      "loss": 0.9561,
+      "step": 12910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4715736755280497,
+      "learning_rate": 1.920199305595915e-05,
+      "loss": 1.3105,
+      "step": 12911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.70138705115768,
+      "learning_rate": 1.91979362965816e-05,
+      "loss": 2.1641,
+      "step": 12912
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.950107915828616,
+      "learning_rate": 1.9193879698673113e-05,
+      "loss": 0.8867,
+      "step": 12913
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.596168141117217,
+      "learning_rate": 1.918982326234659e-05,
+      "loss": 1.2773,
+      "step": 12914
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3540226940140374,
+      "learning_rate": 1.9185766987714927e-05,
+      "loss": 1.2812,
+      "step": 12915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8386968877180663,
+      "learning_rate": 1.9181710874890997e-05,
+      "loss": 0.6501,
+      "step": 12916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8751447889938033,
+      "learning_rate": 1.9177654923987688e-05,
+      "loss": 1.4092,
+      "step": 12917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.248888694977197,
+      "learning_rate": 1.917359913511787e-05,
+      "loss": 1.418,
+      "step": 12918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.554716813531313,
+      "learning_rate": 1.916954350839442e-05,
+      "loss": 1.0645,
+      "step": 12919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.709502002242975,
+      "learning_rate": 1.9165488043930193e-05,
+      "loss": 1.2324,
+      "step": 12920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.307435753817559,
+      "learning_rate": 1.916143274183807e-05,
+      "loss": 1.0469,
+      "step": 12921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.487624423286506,
+      "learning_rate": 1.9157377602230875e-05,
+      "loss": 1.3252,
+      "step": 12922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.809289328825092,
+      "learning_rate": 1.91533226252215e-05,
+      "loss": 0.8755,
+      "step": 12923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.566425046074447,
+      "learning_rate": 1.9149267810922773e-05,
+      "loss": 1.1758,
+      "step": 12924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3479219054666025,
+      "learning_rate": 1.9145213159447542e-05,
+      "loss": 1.1934,
+      "step": 12925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7735100535184243,
+      "learning_rate": 1.9141158670908642e-05,
+      "loss": 0.9941,
+      "step": 12926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2091790277808796,
+      "learning_rate": 1.913710434541892e-05,
+      "loss": 1.582,
+      "step": 12927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7992869236263334,
+      "learning_rate": 1.9133050183091187e-05,
+      "loss": 1.377,
+      "step": 12928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.175772799848208,
+      "learning_rate": 1.912899618403829e-05,
+      "loss": 1.5312,
+      "step": 12929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.492499352917066,
+      "learning_rate": 1.9124942348373025e-05,
+      "loss": 1.7871,
+      "step": 12930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7777793266700248,
+      "learning_rate": 1.9120888676208227e-05,
+      "loss": 0.7368,
+      "step": 12931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3405168535446603,
+      "learning_rate": 1.9116835167656706e-05,
+      "loss": 1.2461,
+      "step": 12932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.276983339045581,
+      "learning_rate": 1.911278182283127e-05,
+      "loss": 0.9912,
+      "step": 12933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.523543246951867,
+      "learning_rate": 1.910872864184472e-05,
+      "loss": 1.2207,
+      "step": 12934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6331815798737592,
+      "learning_rate": 1.9104675624809847e-05,
+      "loss": 1.6816,
+      "step": 12935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.104766190428678,
+      "learning_rate": 1.910062277183946e-05,
+      "loss": 0.8096,
+      "step": 12936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.314186815727424,
+      "learning_rate": 1.9096570083046333e-05,
+      "loss": 1.0205,
+      "step": 12937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.030506947476546,
+      "learning_rate": 1.909251755854326e-05,
+      "loss": 0.6826,
+      "step": 12938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.469805202941869,
+      "learning_rate": 1.9088465198443017e-05,
+      "loss": 1.8027,
+      "step": 12939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.842995008093128,
+      "learning_rate": 1.9084413002858372e-05,
+      "loss": 1.2559,
+      "step": 12940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.169998139823103,
+      "learning_rate": 1.9080360971902116e-05,
+      "loss": 0.9209,
+      "step": 12941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.214009227145453,
+      "learning_rate": 1.9076309105687005e-05,
+      "loss": 0.8262,
+      "step": 12942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.353733144591148,
+      "learning_rate": 1.9072257404325793e-05,
+      "loss": 0.8486,
+      "step": 12943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0531276104821923,
+      "learning_rate": 1.9068205867931254e-05,
+      "loss": 1.21,
+      "step": 12944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.625501675577815,
+      "learning_rate": 1.9064154496616125e-05,
+      "loss": 0.8379,
+      "step": 12945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.81398767369088,
+      "learning_rate": 1.906010329049316e-05,
+      "loss": 1.7129,
+      "step": 12946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3038213621730597,
+      "learning_rate": 1.90560522496751e-05,
+      "loss": 0.6787,
+      "step": 12947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5494094127325346,
+      "learning_rate": 1.9052001374274694e-05,
+      "loss": 1.291,
+      "step": 12948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7907606239297746,
+      "learning_rate": 1.9047950664404663e-05,
+      "loss": 1.1348,
+      "step": 12949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.868074405989458,
+      "learning_rate": 1.9043900120177735e-05,
+      "loss": 0.8936,
+      "step": 12950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7487090288398717,
+      "learning_rate": 1.903984974170665e-05,
+      "loss": 0.9531,
+      "step": 12951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.460958426250314,
+      "learning_rate": 1.903579952910412e-05,
+      "loss": 1.0601,
+      "step": 12952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2345959868931535,
+      "learning_rate": 1.9031749482482862e-05,
+      "loss": 1.1211,
+      "step": 12953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.752407668419302,
+      "learning_rate": 1.9027699601955583e-05,
+      "loss": 1.1992,
+      "step": 12954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.969060941775457,
+      "learning_rate": 1.9023649887634997e-05,
+      "loss": 0.6934,
+      "step": 12955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1704336090022256,
+      "learning_rate": 1.9019600339633798e-05,
+      "loss": 0.7617,
+      "step": 12956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.2957755001730815,
+      "learning_rate": 1.9015550958064694e-05,
+      "loss": 1.4414,
+      "step": 12957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.094941865218646,
+      "learning_rate": 1.901150174304036e-05,
+      "loss": 1.2754,
+      "step": 12958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.209554926834738,
+      "learning_rate": 1.9007452694673495e-05,
+      "loss": 1.1504,
+      "step": 12959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.755536381734826,
+      "learning_rate": 1.9003403813076782e-05,
+      "loss": 1.4424,
+      "step": 12960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6015290033268093,
+      "learning_rate": 1.8999355098362907e-05,
+      "loss": 0.751,
+      "step": 12961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1932694781012563,
+      "learning_rate": 1.8995306550644532e-05,
+      "loss": 1.1992,
+      "step": 12962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.605477882971856,
+      "learning_rate": 1.8991258170034336e-05,
+      "loss": 0.8016,
+      "step": 12963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.32402168528408,
+      "learning_rate": 1.8987209956644975e-05,
+      "loss": 1.2715,
+      "step": 12964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.521012029277908,
+      "learning_rate": 1.8983161910589114e-05,
+      "loss": 1.2549,
+      "step": 12965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6554320852945998,
+      "learning_rate": 1.8979114031979405e-05,
+      "loss": 0.9956,
+      "step": 12966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.858454822001108,
+      "learning_rate": 1.897506632092851e-05,
+      "loss": 0.9648,
+      "step": 12967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.487392193337022,
+      "learning_rate": 1.897101877754905e-05,
+      "loss": 1.0195,
+      "step": 12968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.07923213502264,
+      "learning_rate": 1.8966971401953692e-05,
+      "loss": 1.3457,
+      "step": 12969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.304125041012516,
+      "learning_rate": 1.896292419425507e-05,
+      "loss": 1.3184,
+      "step": 12970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1137750059369504,
+      "learning_rate": 1.8958877154565803e-05,
+      "loss": 0.6641,
+      "step": 12971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.826809075308305,
+      "learning_rate": 1.895483028299853e-05,
+      "loss": 1.3789,
+      "step": 12972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1818643752373834,
+      "learning_rate": 1.8950783579665864e-05,
+      "loss": 1.2734,
+      "step": 12973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.729245415247887,
+      "learning_rate": 1.8946737044680436e-05,
+      "loss": 1.3496,
+      "step": 12974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.931743907189253,
+      "learning_rate": 1.8942690678154845e-05,
+      "loss": 1.3584,
+      "step": 12975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4032859877297916,
+      "learning_rate": 1.8938644480201713e-05,
+      "loss": 0.9771,
+      "step": 12976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1302984042972284,
+      "learning_rate": 1.8934598450933623e-05,
+      "loss": 0.8643,
+      "step": 12977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9059518049297164,
+      "learning_rate": 1.8930552590463198e-05,
+      "loss": 0.8916,
+      "step": 12978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.3770825470323445,
+      "learning_rate": 1.8926506898903025e-05,
+      "loss": 0.9541,
+      "step": 12979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9670829375507086,
+      "learning_rate": 1.8922461376365694e-05,
+      "loss": 1.2578,
+      "step": 12980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9304552788106455,
+      "learning_rate": 1.8918416022963784e-05,
+      "loss": 1.4668,
+      "step": 12981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.524063765192477,
+      "learning_rate": 1.8914370838809885e-05,
+      "loss": 1.3652,
+      "step": 12982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.752822285954488,
+      "learning_rate": 1.8910325824016564e-05,
+      "loss": 1.1299,
+      "step": 12983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.71737594223049,
+      "learning_rate": 1.8906280978696403e-05,
+      "loss": 0.9463,
+      "step": 12984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.73803117728473,
+      "learning_rate": 1.8902236302961952e-05,
+      "loss": 1.3994,
+      "step": 12985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.453603291782025,
+      "learning_rate": 1.8898191796925792e-05,
+      "loss": 0.8301,
+      "step": 12986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1657083634059964,
+      "learning_rate": 1.889414746070046e-05,
+      "loss": 0.7236,
+      "step": 12987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1046981392119273,
+      "learning_rate": 1.889010329439852e-05,
+      "loss": 1.4395,
+      "step": 12988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.745017296590041,
+      "learning_rate": 1.8886059298132524e-05,
+      "loss": 1.2695,
+      "step": 12989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0526969753818265,
+      "learning_rate": 1.888201547201501e-05,
+      "loss": 0.5645,
+      "step": 12990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1272018305119555,
+      "learning_rate": 1.887797181615851e-05,
+      "loss": 1.1543,
+      "step": 12991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.025380537830182,
+      "learning_rate": 1.887392833067556e-05,
+      "loss": 0.7261,
+      "step": 12992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6895995699281365,
+      "learning_rate": 1.88698850156787e-05,
+      "loss": 0.6943,
+      "step": 12993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6581034815743951,
+      "learning_rate": 1.8865841871280438e-05,
+      "loss": 0.9893,
+      "step": 12994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2283559971039417,
+      "learning_rate": 1.8861798897593303e-05,
+      "loss": 0.9287,
+      "step": 12995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8375643012873613,
+      "learning_rate": 1.8857756094729794e-05,
+      "loss": 1.2266,
+      "step": 12996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4154962029116334,
+      "learning_rate": 1.885371346280244e-05,
+      "loss": 0.7197,
+      "step": 12997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.246639862211,
+      "learning_rate": 1.884967100192374e-05,
+      "loss": 0.9561,
+      "step": 12998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8909487981124995,
+      "learning_rate": 1.884562871220619e-05,
+      "loss": 1.1865,
+      "step": 12999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2784395806983464,
+      "learning_rate": 1.884158659376229e-05,
+      "loss": 0.958,
+      "step": 13000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.377617841677023,
+      "learning_rate": 1.883754464670453e-05,
+      "loss": 1.1641,
+      "step": 13001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2910709802068638,
+      "learning_rate": 1.8833502871145387e-05,
+      "loss": 0.9712,
+      "step": 13002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8271265954737004,
+      "learning_rate": 1.8829461267197356e-05,
+      "loss": 0.8936,
+      "step": 13003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.705984079398379,
+      "learning_rate": 1.8825419834972902e-05,
+      "loss": 0.7109,
+      "step": 13004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.525936348646213,
+      "learning_rate": 1.882137857458449e-05,
+      "loss": 1.3262,
+      "step": 13005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.039086074027534,
+      "learning_rate": 1.881733748614461e-05,
+      "loss": 1.3975,
+      "step": 13006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.024839997517322,
+      "learning_rate": 1.8813296569765708e-05,
+      "loss": 0.7676,
+      "step": 13007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.222023605114442,
+      "learning_rate": 1.880925582556025e-05,
+      "loss": 0.666,
+      "step": 13008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2185804275401773,
+      "learning_rate": 1.8805215253640674e-05,
+      "loss": 1.0938,
+      "step": 13009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6215655401172495,
+      "learning_rate": 1.880117485411944e-05,
+      "loss": 1.5684,
+      "step": 13010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7991902569839038,
+      "learning_rate": 1.8797134627108984e-05,
+      "loss": 1.332,
+      "step": 13011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.105376296086718,
+      "learning_rate": 1.879309457272175e-05,
+      "loss": 0.7686,
+      "step": 13012
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6648670223475794,
+      "learning_rate": 1.8789054691070163e-05,
+      "loss": 0.9375,
+      "step": 13013
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.708859925550312,
+      "learning_rate": 1.878501498226666e-05,
+      "loss": 1.0322,
+      "step": 13014
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.867742749742161,
+      "learning_rate": 1.8780975446423646e-05,
+      "loss": 1.1367,
+      "step": 13015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.732404883489254,
+      "learning_rate": 1.877693608365356e-05,
+      "loss": 1.2871,
+      "step": 13016
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.220038352431027,
+      "learning_rate": 1.8772896894068816e-05,
+      "loss": 2.084,
+      "step": 13017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.465223151833716,
+      "learning_rate": 1.8768857877781813e-05,
+      "loss": 0.7012,
+      "step": 13018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.403238864775662,
+      "learning_rate": 1.8764819034904956e-05,
+      "loss": 0.9658,
+      "step": 13019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.302771190796802,
+      "learning_rate": 1.8760780365550655e-05,
+      "loss": 0.7803,
+      "step": 13020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5163166637319048,
+      "learning_rate": 1.875674186983129e-05,
+      "loss": 1.0356,
+      "step": 13021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3687550607591987,
+      "learning_rate": 1.875270354785926e-05,
+      "loss": 0.9114,
+      "step": 13022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4911935193310124,
+      "learning_rate": 1.8748665399746947e-05,
+      "loss": 1.29,
+      "step": 13023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.257299239482303,
+      "learning_rate": 1.874462742560672e-05,
+      "loss": 1.208,
+      "step": 13024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.004490460459465,
+      "learning_rate": 1.8740589625550977e-05,
+      "loss": 1.1504,
+      "step": 13025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6364010767834234,
+      "learning_rate": 1.8736551999692075e-05,
+      "loss": 0.9824,
+      "step": 13026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.981994875419381,
+      "learning_rate": 1.873251454814238e-05,
+      "loss": 1.8281,
+      "step": 13027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7101452530411045,
+      "learning_rate": 1.8728477271014252e-05,
+      "loss": 0.4301,
+      "step": 13028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.284464309136393,
+      "learning_rate": 1.8724440168420053e-05,
+      "loss": 1.0342,
+      "step": 13029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7064089369409525,
+      "learning_rate": 1.8720403240472125e-05,
+      "loss": 1.0352,
+      "step": 13030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.574471014740094,
+      "learning_rate": 1.8716366487282824e-05,
+      "loss": 0.7607,
+      "step": 13031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4404722368661087,
+      "learning_rate": 1.871232990896448e-05,
+      "loss": 0.6211,
+      "step": 13032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5313278939777506,
+      "learning_rate": 1.870829350562943e-05,
+      "loss": 1.6855,
+      "step": 13033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.997172429140841,
+      "learning_rate": 1.8704257277390018e-05,
+      "loss": 1.1992,
+      "step": 13034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.419481256346192,
+      "learning_rate": 1.8700221224358562e-05,
+      "loss": 0.6626,
+      "step": 13035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9256203865678247,
+      "learning_rate": 1.8696185346647388e-05,
+      "loss": 1.3857,
+      "step": 13036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7241202934987503,
+      "learning_rate": 1.86921496443688e-05,
+      "loss": 0.8198,
+      "step": 13037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6895632253811685,
+      "learning_rate": 1.868811411763513e-05,
+      "loss": 1.3477,
+      "step": 13038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6768131243430178,
+      "learning_rate": 1.8684078766558667e-05,
+      "loss": 0.9336,
+      "step": 13039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9789624627491895,
+      "learning_rate": 1.8680043591251723e-05,
+      "loss": 0.9629,
+      "step": 13040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4857746715915394,
+      "learning_rate": 1.86760085918266e-05,
+      "loss": 1.0576,
+      "step": 13041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2068287347561326,
+      "learning_rate": 1.8671973768395567e-05,
+      "loss": 0.7207,
+      "step": 13042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.750530625214953,
+      "learning_rate": 1.8667939121070937e-05,
+      "loss": 0.9414,
+      "step": 13043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.651575890295401,
+      "learning_rate": 1.8663904649964986e-05,
+      "loss": 1.4512,
+      "step": 13044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6646526499648315,
+      "learning_rate": 1.8659870355189984e-05,
+      "loss": 1.0234,
+      "step": 13045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4744218786876315,
+      "learning_rate": 1.865583623685822e-05,
+      "loss": 0.8184,
+      "step": 13046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1646986484254276,
+      "learning_rate": 1.8651802295081937e-05,
+      "loss": 0.8262,
+      "step": 13047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.910052331992884,
+      "learning_rate": 1.8647768529973425e-05,
+      "loss": 1.084,
+      "step": 13048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.801715604531323,
+      "learning_rate": 1.864373494164492e-05,
+      "loss": 0.9258,
+      "step": 13049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.901440290349632,
+      "learning_rate": 1.863970153020869e-05,
+      "loss": 0.6289,
+      "step": 13050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.5461734811672745,
+      "learning_rate": 1.863566829577697e-05,
+      "loss": 1.2734,
+      "step": 13051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8479147795024686,
+      "learning_rate": 1.8631635238462006e-05,
+      "loss": 0.9189,
+      "step": 13052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8651528225790794,
+      "learning_rate": 1.8627602358376052e-05,
+      "loss": 0.594,
+      "step": 13053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9375214271575563,
+      "learning_rate": 1.8623569655631322e-05,
+      "loss": 0.8242,
+      "step": 13054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.061912568563267,
+      "learning_rate": 1.8619537130340063e-05,
+      "loss": 0.9902,
+      "step": 13055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3392771814760396,
+      "learning_rate": 1.8615504782614478e-05,
+      "loss": 1.5762,
+      "step": 13056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.069935083852325,
+      "learning_rate": 1.8611472612566804e-05,
+      "loss": 1.0225,
+      "step": 13057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.249424118942963,
+      "learning_rate": 1.860744062030924e-05,
+      "loss": 0.7002,
+      "step": 13058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.335561358427192,
+      "learning_rate": 1.8603408805954005e-05,
+      "loss": 1.4785,
+      "step": 13059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.662479668778893,
+      "learning_rate": 1.8599377169613297e-05,
+      "loss": 0.9531,
+      "step": 13060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1206317208720726,
+      "learning_rate": 1.8595345711399307e-05,
+      "loss": 0.8359,
+      "step": 13061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0183861324839376,
+      "learning_rate": 1.859131443142425e-05,
+      "loss": 0.6431,
+      "step": 13062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.385054016824507,
+      "learning_rate": 1.8587283329800302e-05,
+      "loss": 1.2637,
+      "step": 13063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7472494171093573,
+      "learning_rate": 1.858325240663965e-05,
+      "loss": 0.8538,
+      "step": 13064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2240408555264404,
+      "learning_rate": 1.8579221662054474e-05,
+      "loss": 1.291,
+      "step": 13065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.222432738766245,
+      "learning_rate": 1.8575191096156935e-05,
+      "loss": 0.9922,
+      "step": 13066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4622331387526106,
+      "learning_rate": 1.8571160709059225e-05,
+      "loss": 0.9307,
+      "step": 13067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.250423603448581,
+      "learning_rate": 1.8567130500873485e-05,
+      "loss": 0.9814,
+      "step": 13068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.007933775203221,
+      "learning_rate": 1.856310047171189e-05,
+      "loss": 1.6211,
+      "step": 13069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2557658386655253,
+      "learning_rate": 1.8559070621686582e-05,
+      "loss": 0.8975,
+      "step": 13070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.041458067325757,
+      "learning_rate": 1.8555040950909718e-05,
+      "loss": 1.0127,
+      "step": 13071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.213853941189143,
+      "learning_rate": 1.8551011459493445e-05,
+      "loss": 1.6797,
+      "step": 13072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1194407729548557,
+      "learning_rate": 1.8546982147549895e-05,
+      "loss": 0.7402,
+      "step": 13073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.448446607027691,
+      "learning_rate": 1.8542953015191207e-05,
+      "loss": 0.8787,
+      "step": 13074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.7617660157108626,
+      "learning_rate": 1.8538924062529506e-05,
+      "loss": 0.7051,
+      "step": 13075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.792806248167726,
+      "learning_rate": 1.853489528967692e-05,
+      "loss": 1.3855,
+      "step": 13076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.197107416179277,
+      "learning_rate": 1.8530866696745565e-05,
+      "loss": 0.835,
+      "step": 13077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.707470719625363,
+      "learning_rate": 1.852683828384756e-05,
+      "loss": 1.623,
+      "step": 13078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.142251281011374,
+      "learning_rate": 1.852281005109499e-05,
+      "loss": 0.8818,
+      "step": 13079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7152932652527433,
+      "learning_rate": 1.8518781998599997e-05,
+      "loss": 1.0488,
+      "step": 13080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.746622503395643,
+      "learning_rate": 1.8514754126474665e-05,
+      "loss": 1.0391,
+      "step": 13081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7455974786748956,
+      "learning_rate": 1.8510726434831082e-05,
+      "loss": 0.4406,
+      "step": 13082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.159001543944485,
+      "learning_rate": 1.850669892378134e-05,
+      "loss": 1.293,
+      "step": 13083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5182755054101373,
+      "learning_rate": 1.8502671593437526e-05,
+      "loss": 1.166,
+      "step": 13084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5923195023610384,
+      "learning_rate": 1.8498644443911712e-05,
+      "loss": 0.54,
+      "step": 13085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3409798082004776,
+      "learning_rate": 1.8494617475315985e-05,
+      "loss": 1.3789,
+      "step": 13086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1850922548908622,
+      "learning_rate": 1.8490590687762393e-05,
+      "loss": 1.165,
+      "step": 13087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.299822763738122,
+      "learning_rate": 1.8486564081363022e-05,
+      "loss": 1.5039,
+      "step": 13088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4945034736400986,
+      "learning_rate": 1.8482537656229912e-05,
+      "loss": 1.0049,
+      "step": 13089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.479373143250463,
+      "learning_rate": 1.847851141247513e-05,
+      "loss": 0.8877,
+      "step": 13090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.4300366891413745,
+      "learning_rate": 1.8474485350210725e-05,
+      "loss": 1.2656,
+      "step": 13091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4830665261570286,
+      "learning_rate": 1.847045946954873e-05,
+      "loss": 1.3574,
+      "step": 13092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.595904058074339,
+      "learning_rate": 1.8466433770601194e-05,
+      "loss": 1.2529,
+      "step": 13093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9173803728330396,
+      "learning_rate": 1.8462408253480142e-05,
+      "loss": 1.5176,
+      "step": 13094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.813748866409882,
+      "learning_rate": 1.8458382918297614e-05,
+      "loss": 1.4131,
+      "step": 13095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3996987994798977,
+      "learning_rate": 1.845435776516562e-05,
+      "loss": 1.4385,
+      "step": 13096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.233779778743393,
+      "learning_rate": 1.845033279419619e-05,
+      "loss": 1.4336,
+      "step": 13097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5479818174111664,
+      "learning_rate": 1.8446308005501318e-05,
+      "loss": 1.3857,
+      "step": 13098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1420902333064666,
+      "learning_rate": 1.844228339919304e-05,
+      "loss": 1.1953,
+      "step": 13099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.924293713495446,
+      "learning_rate": 1.843825897538334e-05,
+      "loss": 1.2793,
+      "step": 13100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.93330749659286,
+      "learning_rate": 1.8434234734184225e-05,
+      "loss": 0.6084,
+      "step": 13101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.943973608675359,
+      "learning_rate": 1.843021067570768e-05,
+      "loss": 1.2246,
+      "step": 13102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2870520398049274,
+      "learning_rate": 1.8426186800065704e-05,
+      "loss": 0.6543,
+      "step": 13103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4236428906502785,
+      "learning_rate": 1.8422163107370266e-05,
+      "loss": 1.3154,
+      "step": 13104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4414328123555014,
+      "learning_rate": 1.841813959773336e-05,
+      "loss": 1.002,
+      "step": 13105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7816544142312867,
+      "learning_rate": 1.841411627126694e-05,
+      "loss": 1.0728,
+      "step": 13106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.882400784252634,
+      "learning_rate": 1.8410093128082982e-05,
+      "loss": 0.6149,
+      "step": 13107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7052183479845424,
+      "learning_rate": 1.8406070168293457e-05,
+      "loss": 0.811,
+      "step": 13108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5174871151546103,
+      "learning_rate": 1.840204739201031e-05,
+      "loss": 0.7754,
+      "step": 13109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8734349394741807,
+      "learning_rate": 1.8398024799345502e-05,
+      "loss": 1.3447,
+      "step": 13110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40286432844137,
+      "learning_rate": 1.8394002390410976e-05,
+      "loss": 1.1035,
+      "step": 13111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.068915597853567,
+      "learning_rate": 1.8389980165318673e-05,
+      "loss": 1.2007,
+      "step": 13112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.009824560381775,
+      "learning_rate": 1.838595812418053e-05,
+      "loss": 1.3477,
+      "step": 13113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.734620465837912,
+      "learning_rate": 1.8381936267108485e-05,
+      "loss": 1.5605,
+      "step": 13114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.294151819774374,
+      "learning_rate": 1.837791459421445e-05,
+      "loss": 0.6875,
+      "step": 13115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.471231493034007,
+      "learning_rate": 1.8373893105610356e-05,
+      "loss": 0.7217,
+      "step": 13116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.020602530946993,
+      "learning_rate": 1.8369871801408123e-05,
+      "loss": 0.5183,
+      "step": 13117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.543866680196702,
+      "learning_rate": 1.836585068171966e-05,
+      "loss": 1.2559,
+      "step": 13118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.964723602128145,
+      "learning_rate": 1.8361829746656873e-05,
+      "loss": 1.1201,
+      "step": 13119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8102121712480812,
+      "learning_rate": 1.835780899633166e-05,
+      "loss": 0.467,
+      "step": 13120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.811518433333251,
+      "learning_rate": 1.835378843085592e-05,
+      "loss": 0.707,
+      "step": 13121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.226531561148825,
+      "learning_rate": 1.8349768050341544e-05,
+      "loss": 1.1953,
+      "step": 13122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1447897858101794,
+      "learning_rate": 1.834574785490041e-05,
+      "loss": 0.998,
+      "step": 13123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9108495409755473,
+      "learning_rate": 1.8341727844644414e-05,
+      "loss": 1.3574,
+      "step": 13124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.563501481433317,
+      "learning_rate": 1.8337708019685412e-05,
+      "loss": 1.4199,
+      "step": 13125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3735546934465157,
+      "learning_rate": 1.833368838013528e-05,
+      "loss": 0.5958,
+      "step": 13126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.510173529693102,
+      "learning_rate": 1.8329668926105896e-05,
+      "loss": 1.1636,
+      "step": 13127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.027494627055282,
+      "learning_rate": 1.8325649657709103e-05,
+      "loss": 1.0791,
+      "step": 13128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1726048841717747,
+      "learning_rate": 1.832163057505677e-05,
+      "loss": 0.8816,
+      "step": 13129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9744267849102464,
+      "learning_rate": 1.8317611678260737e-05,
+      "loss": 1.2002,
+      "step": 13130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.864303395659487,
+      "learning_rate": 1.831359296743285e-05,
+      "loss": 1.4375,
+      "step": 13131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4296079174881173,
+      "learning_rate": 1.8309574442684952e-05,
+      "loss": 0.7939,
+      "step": 13132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1825911893725674,
+      "learning_rate": 1.830555610412887e-05,
+      "loss": 0.8438,
+      "step": 13133
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.204178253989479,
+      "learning_rate": 1.8301537951876434e-05,
+      "loss": 1.3867,
+      "step": 13134
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6154703303351408,
+      "learning_rate": 1.8297519986039468e-05,
+      "loss": 0.4575,
+      "step": 13135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5241215490282554,
+      "learning_rate": 1.8293502206729794e-05,
+      "loss": 0.2957,
+      "step": 13136
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8434918324679135,
+      "learning_rate": 1.828948461405923e-05,
+      "loss": 1.2891,
+      "step": 13137
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.435894755440497,
+      "learning_rate": 1.8285467208139573e-05,
+      "loss": 1.0498,
+      "step": 13138
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.281536972599047,
+      "learning_rate": 1.8281449989082634e-05,
+      "loss": 0.8882,
+      "step": 13139
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9622947782747002,
+      "learning_rate": 1.8277432957000202e-05,
+      "loss": 1.0957,
+      "step": 13140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.381374137626365,
+      "learning_rate": 1.8273416112004075e-05,
+      "loss": 1.0918,
+      "step": 13141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.459682471554075,
+      "learning_rate": 1.8269399454206038e-05,
+      "loss": 0.6924,
+      "step": 13142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.627177042709809,
+      "learning_rate": 1.826538298371788e-05,
+      "loss": 1.165,
+      "step": 13143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.493418649227619,
+      "learning_rate": 1.826136670065136e-05,
+      "loss": 0.8848,
+      "step": 13144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.142472667833185,
+      "learning_rate": 1.8257350605118266e-05,
+      "loss": 0.8857,
+      "step": 13145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.6659952229027875,
+      "learning_rate": 1.8253334697230366e-05,
+      "loss": 1.5059,
+      "step": 13146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1580464775713555,
+      "learning_rate": 1.824931897709941e-05,
+      "loss": 1.2012,
+      "step": 13147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6816847171829183,
+      "learning_rate": 1.8245303444837165e-05,
+      "loss": 0.4353,
+      "step": 13148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.815623895268511,
+      "learning_rate": 1.824128810055537e-05,
+      "loss": 1.0879,
+      "step": 13149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.406047862096111,
+      "learning_rate": 1.823727294436578e-05,
+      "loss": 0.9883,
+      "step": 13150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1400791282159624,
+      "learning_rate": 1.8233257976380126e-05,
+      "loss": 1.1133,
+      "step": 13151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.955206881764763,
+      "learning_rate": 1.822924319671015e-05,
+      "loss": 0.8857,
+      "step": 13152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1534499633233537,
+      "learning_rate": 1.8225228605467576e-05,
+      "loss": 1.0198,
+      "step": 13153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.803011337578525,
+      "learning_rate": 1.8221214202764125e-05,
+      "loss": 1.2207,
+      "step": 13154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.112956069562793,
+      "learning_rate": 1.8217199988711537e-05,
+      "loss": 1.1506,
+      "step": 13155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4851535560859532,
+      "learning_rate": 1.8213185963421504e-05,
+      "loss": 0.8823,
+      "step": 13156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3892830466842665,
+      "learning_rate": 1.8209172127005748e-05,
+      "loss": 1.0332,
+      "step": 13157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.402282217834944,
+      "learning_rate": 1.820515847957597e-05,
+      "loss": 0.877,
+      "step": 13158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.082326896614831,
+      "learning_rate": 1.8201145021243855e-05,
+      "loss": 1.0684,
+      "step": 13159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2373404930842846,
+      "learning_rate": 1.8197131752121117e-05,
+      "loss": 0.9209,
+      "step": 13160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.947818570251588,
+      "learning_rate": 1.8193118672319425e-05,
+      "loss": 0.8419,
+      "step": 13161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9414167872590973,
+      "learning_rate": 1.8189105781950475e-05,
+      "loss": 0.9492,
+      "step": 13162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.531692918188581,
+      "learning_rate": 1.8185093081125926e-05,
+      "loss": 1.0659,
+      "step": 13163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4610321511125806,
+      "learning_rate": 1.8181080569957472e-05,
+      "loss": 1.3594,
+      "step": 13164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.653213740326148,
+      "learning_rate": 1.8177068248556772e-05,
+      "loss": 1.29,
+      "step": 13165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8631422842734673,
+      "learning_rate": 1.817305611703548e-05,
+      "loss": 0.6421,
+      "step": 13166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.234825565157183,
+      "learning_rate": 1.8169044175505263e-05,
+      "loss": 1.1992,
+      "step": 13167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8806953414707768,
+      "learning_rate": 1.816503242407776e-05,
+      "loss": 1.3418,
+      "step": 13168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9475281979520007,
+      "learning_rate": 1.816102086286463e-05,
+      "loss": 0.5823,
+      "step": 13169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7852329095787112,
+      "learning_rate": 1.8157009491977495e-05,
+      "loss": 0.6108,
+      "step": 13170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.732777552317503,
+      "learning_rate": 1.8152998311528006e-05,
+      "loss": 1.8438,
+      "step": 13171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0037911620000597,
+      "learning_rate": 1.814898732162778e-05,
+      "loss": 0.8018,
+      "step": 13172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.271112255733589,
+      "learning_rate": 1.8144976522388452e-05,
+      "loss": 1.4688,
+      "step": 13173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.535491556525719,
+      "learning_rate": 1.814096591392164e-05,
+      "loss": 1.0762,
+      "step": 13174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.565194967751561,
+      "learning_rate": 1.813695549633895e-05,
+      "loss": 1.4785,
+      "step": 13175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.771322664718369,
+      "learning_rate": 1.8132945269751994e-05,
+      "loss": 1.0029,
+      "step": 13176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.525573921601804,
+      "learning_rate": 1.812893523427238e-05,
+      "loss": 1.0,
+      "step": 13177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1453149802155056,
+      "learning_rate": 1.8124925390011697e-05,
+      "loss": 1.5605,
+      "step": 13178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.514558080142398,
+      "learning_rate": 1.8120915737081547e-05,
+      "loss": 1.1895,
+      "step": 13179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.346705224107566,
+      "learning_rate": 1.811690627559351e-05,
+      "loss": 1.416,
+      "step": 13180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.560318128108757,
+      "learning_rate": 1.811289700565916e-05,
+      "loss": 1.1338,
+      "step": 13181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.268255012764341,
+      "learning_rate": 1.8108887927390092e-05,
+      "loss": 1.2627,
+      "step": 13182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.854732861634552,
+      "learning_rate": 1.8104879040897864e-05,
+      "loss": 1.4746,
+      "step": 13183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.809611638746073,
+      "learning_rate": 1.8100870346294048e-05,
+      "loss": 1.1523,
+      "step": 13184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.091140065288311,
+      "learning_rate": 1.80968618436902e-05,
+      "loss": 1.3662,
+      "step": 13185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.393493754602152,
+      "learning_rate": 1.809285353319788e-05,
+      "loss": 1.1025,
+      "step": 13186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.272150088800384,
+      "learning_rate": 1.808884541492863e-05,
+      "loss": 1.2793,
+      "step": 13187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4070264298847324,
+      "learning_rate": 1.8084837488994006e-05,
+      "loss": 1.5527,
+      "step": 13188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.482057366070159,
+      "learning_rate": 1.808082975550553e-05,
+      "loss": 1.251,
+      "step": 13189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4525807742816994,
+      "learning_rate": 1.807682221457475e-05,
+      "loss": 1.1133,
+      "step": 13190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6491468711738215,
+      "learning_rate": 1.807281486631318e-05,
+      "loss": 1.3691,
+      "step": 13191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7723470773995667,
+      "learning_rate": 1.806880771083236e-05,
+      "loss": 1.1719,
+      "step": 13192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7073290175703373,
+      "learning_rate": 1.8064800748243803e-05,
+      "loss": 1.4961,
+      "step": 13193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.004944606212192,
+      "learning_rate": 1.8060793978659014e-05,
+      "loss": 1.0996,
+      "step": 13194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.594220589840039,
+      "learning_rate": 1.8056787402189506e-05,
+      "loss": 1.1992,
+      "step": 13195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.386409758250087,
+      "learning_rate": 1.8052781018946778e-05,
+      "loss": 0.6807,
+      "step": 13196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.856188556836644,
+      "learning_rate": 1.804877482904232e-05,
+      "loss": 1.0273,
+      "step": 13197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1404354262116243,
+      "learning_rate": 1.804476883258764e-05,
+      "loss": 0.8696,
+      "step": 13198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8278392340168697,
+      "learning_rate": 1.80407630296942e-05,
+      "loss": 0.8787,
+      "step": 13199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2641601282938995,
+      "learning_rate": 1.803675742047349e-05,
+      "loss": 1.1699,
+      "step": 13200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.233342681834394,
+      "learning_rate": 1.8032752005036997e-05,
+      "loss": 1.4863,
+      "step": 13201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.650633542277899,
+      "learning_rate": 1.8028746783496174e-05,
+      "loss": 1.5801,
+      "step": 13202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.995754537631783,
+      "learning_rate": 1.802474175596249e-05,
+      "loss": 0.9941,
+      "step": 13203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1209528847083443,
+      "learning_rate": 1.802073692254741e-05,
+      "loss": 1.0547,
+      "step": 13204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3764813315870934,
+      "learning_rate": 1.8016732283362374e-05,
+      "loss": 0.9878,
+      "step": 13205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6665593860822177,
+      "learning_rate": 1.8012727838518838e-05,
+      "loss": 1.0259,
+      "step": 13206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5249739655011774,
+      "learning_rate": 1.800872358812824e-05,
+      "loss": 1.2493,
+      "step": 13207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4594120202192586,
+      "learning_rate": 1.800471953230202e-05,
+      "loss": 1.4922,
+      "step": 13208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5220236581998057,
+      "learning_rate": 1.8000715671151597e-05,
+      "loss": 0.6609,
+      "step": 13209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.944672235218318,
+      "learning_rate": 1.7996712004788422e-05,
+      "loss": 1.2173,
+      "step": 13210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3161132107909443,
+      "learning_rate": 1.7992708533323894e-05,
+      "loss": 0.7432,
+      "step": 13211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.727139015893804,
+      "learning_rate": 1.798870525686944e-05,
+      "loss": 1.6875,
+      "step": 13212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4611801165754783,
+      "learning_rate": 1.7984702175536458e-05,
+      "loss": 1.085,
+      "step": 13213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1497084309018804,
+      "learning_rate": 1.7980699289436365e-05,
+      "loss": 0.9971,
+      "step": 13214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.039132537223614,
+      "learning_rate": 1.797669659868055e-05,
+      "loss": 0.96,
+      "step": 13215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.845256867302173,
+      "learning_rate": 1.7972694103380407e-05,
+      "loss": 1.0029,
+      "step": 13216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.798769346632883,
+      "learning_rate": 1.7968691803647332e-05,
+      "loss": 0.9727,
+      "step": 13217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.395903799154095,
+      "learning_rate": 1.796468969959269e-05,
+      "loss": 1.3906,
+      "step": 13218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.039456267355365,
+      "learning_rate": 1.7960687791327873e-05,
+      "loss": 0.9497,
+      "step": 13219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.366845537419827,
+      "learning_rate": 1.795668607896426e-05,
+      "loss": 1.1582,
+      "step": 13220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.881221575794167,
+      "learning_rate": 1.7952684562613197e-05,
+      "loss": 1.4453,
+      "step": 13221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8784765701864425,
+      "learning_rate": 1.794868324238606e-05,
+      "loss": 1.1128,
+      "step": 13222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9263805516961083,
+      "learning_rate": 1.794468211839419e-05,
+      "loss": 1.2861,
+      "step": 13223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.874778938171958,
+      "learning_rate": 1.7940681190748954e-05,
+      "loss": 1.0742,
+      "step": 13224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.875574850335442,
+      "learning_rate": 1.793668045956168e-05,
+      "loss": 0.8147,
+      "step": 13225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.155171784462603,
+      "learning_rate": 1.793267992494372e-05,
+      "loss": 0.8408,
+      "step": 13226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6996870426009738,
+      "learning_rate": 1.792867958700639e-05,
+      "loss": 0.507,
+      "step": 13227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.7118234892525495,
+      "learning_rate": 1.7924679445861027e-05,
+      "loss": 1.1982,
+      "step": 13228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.83967096571812,
+      "learning_rate": 1.7920679501618965e-05,
+      "loss": 1.04,
+      "step": 13229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.72904666659365,
+      "learning_rate": 1.7916679754391504e-05,
+      "loss": 0.9316,
+      "step": 13230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3299766192265916,
+      "learning_rate": 1.7912680204289965e-05,
+      "loss": 1.2051,
+      "step": 13231
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.774184920336507,
+      "learning_rate": 1.7908680851425652e-05,
+      "loss": 1.4482,
+      "step": 13232
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.87924368044695,
+      "learning_rate": 1.7904681695909865e-05,
+      "loss": 0.8066,
+      "step": 13233
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.738923654557013,
+      "learning_rate": 1.7900682737853894e-05,
+      "loss": 1.1758,
+      "step": 13234
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6534143009324,
+      "learning_rate": 1.7896683977369035e-05,
+      "loss": 1.0352,
+      "step": 13235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.519427642030332,
+      "learning_rate": 1.789268541456657e-05,
+      "loss": 1.1348,
+      "step": 13236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2632352855423394,
+      "learning_rate": 1.788868704955776e-05,
+      "loss": 1.1255,
+      "step": 13237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0387036423028535,
+      "learning_rate": 1.788468888245391e-05,
+      "loss": 1.2404,
+      "step": 13238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3193784946753517,
+      "learning_rate": 1.7880690913366275e-05,
+      "loss": 1.4238,
+      "step": 13239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.257384923312848,
+      "learning_rate": 1.7876693142406105e-05,
+      "loss": 1.8691,
+      "step": 13240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9662949528681612,
+      "learning_rate": 1.7872695569684672e-05,
+      "loss": 0.9951,
+      "step": 13241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.238415991969922,
+      "learning_rate": 1.7868698195313216e-05,
+      "loss": 1.1699,
+      "step": 13242
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.685288554763754,
+      "learning_rate": 1.7864701019402992e-05,
+      "loss": 1.1787,
+      "step": 13243
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.598116090866237,
+      "learning_rate": 1.7860704042065222e-05,
+      "loss": 1.2461,
+      "step": 13244
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.131296523297963,
+      "learning_rate": 1.7856707263411167e-05,
+      "loss": 1.377,
+      "step": 13245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.238689924549085,
+      "learning_rate": 1.785271068355202e-05,
+      "loss": 1.252,
+      "step": 13246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9474758146969293,
+      "learning_rate": 1.784871430259904e-05,
+      "loss": 0.8936,
+      "step": 13247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5122654440032433,
+      "learning_rate": 1.784471812066343e-05,
+      "loss": 1.3223,
+      "step": 13248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.884850089477324,
+      "learning_rate": 1.7840722137856398e-05,
+      "loss": 1.0518,
+      "step": 13249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.817861067509996,
+      "learning_rate": 1.7836726354289158e-05,
+      "loss": 1.1934,
+      "step": 13250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.858448394058487,
+      "learning_rate": 1.78327307700729e-05,
+      "loss": 1.4072,
+      "step": 13251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.254016997381847,
+      "learning_rate": 1.7828735385318835e-05,
+      "loss": 1.6953,
+      "step": 13252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0974888228617643,
+      "learning_rate": 1.782474020013814e-05,
+      "loss": 0.7925,
+      "step": 13253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5935110878355294,
+      "learning_rate": 1.7820745214642008e-05,
+      "loss": 0.8213,
+      "step": 13254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.547986633602455,
+      "learning_rate": 1.7816750428941596e-05,
+      "loss": 0.541,
+      "step": 13255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4919630565446,
+      "learning_rate": 1.781275584314812e-05,
+      "loss": 1.334,
+      "step": 13256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.522046062808137,
+      "learning_rate": 1.7808761457372707e-05,
+      "loss": 0.9277,
+      "step": 13257
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0288343734418963,
+      "learning_rate": 1.7804767271726546e-05,
+      "loss": 1.1133,
+      "step": 13258
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3452454945588395,
+      "learning_rate": 1.7800773286320775e-05,
+      "loss": 1.2617,
+      "step": 13259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5601450636115564,
+      "learning_rate": 1.7796779501266564e-05,
+      "loss": 1.0479,
+      "step": 13260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.156691678320412,
+      "learning_rate": 1.7792785916675036e-05,
+      "loss": 1.0791,
+      "step": 13261
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1535430723780293,
+      "learning_rate": 1.778879253265735e-05,
+      "loss": 1.2773,
+      "step": 13262
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9256548049832634,
+      "learning_rate": 1.7784799349324624e-05,
+      "loss": 1.1797,
+      "step": 13263
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4849487653384075,
+      "learning_rate": 1.7780806366788e-05,
+      "loss": 1.0996,
+      "step": 13264
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7652300536489314,
+      "learning_rate": 1.7776813585158584e-05,
+      "loss": 0.8535,
+      "step": 13265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4851499104711268,
+      "learning_rate": 1.7772821004547518e-05,
+      "loss": 0.9448,
+      "step": 13266
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.899983202129253,
+      "learning_rate": 1.7768828625065902e-05,
+      "loss": 1.124,
+      "step": 13267
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.801346598112477,
+      "learning_rate": 1.7764836446824833e-05,
+      "loss": 1.2012,
+      "step": 13268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.901534818764785,
+      "learning_rate": 1.776084446993543e-05,
+      "loss": 0.7686,
+      "step": 13269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.797596678416678,
+      "learning_rate": 1.7756852694508773e-05,
+      "loss": 1.3418,
+      "step": 13270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.809753554376,
+      "learning_rate": 1.775286112065596e-05,
+      "loss": 0.561,
+      "step": 13271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.090394174810178,
+      "learning_rate": 1.7748869748488068e-05,
+      "loss": 1.1455,
+      "step": 13272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7825586047859985,
+      "learning_rate": 1.7744878578116182e-05,
+      "loss": 0.6936,
+      "step": 13273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.068074063515095,
+      "learning_rate": 1.7740887609651357e-05,
+      "loss": 0.9976,
+      "step": 13274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.083673787358939,
+      "learning_rate": 1.7736896843204692e-05,
+      "loss": 1.0645,
+      "step": 13275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.914961355707999,
+      "learning_rate": 1.7732906278887225e-05,
+      "loss": 0.6709,
+      "step": 13276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5999645560856424,
+      "learning_rate": 1.772891591681002e-05,
+      "loss": 0.8428,
+      "step": 13277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4637390233460987,
+      "learning_rate": 1.7724925757084125e-05,
+      "loss": 0.9141,
+      "step": 13278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.077768969332995,
+      "learning_rate": 1.772093579982059e-05,
+      "loss": 1.3965,
+      "step": 13279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.301760971594652,
+      "learning_rate": 1.7716946045130433e-05,
+      "loss": 1.625,
+      "step": 13280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8837863614209867,
+      "learning_rate": 1.7712956493124717e-05,
+      "loss": 1.165,
+      "step": 13281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.214562367840425,
+      "learning_rate": 1.7708967143914448e-05,
+      "loss": 1.2363,
+      "step": 13282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3746323551924116,
+      "learning_rate": 1.770497799761065e-05,
+      "loss": 1.2988,
+      "step": 13283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9124587833725346,
+      "learning_rate": 1.7700989054324355e-05,
+      "loss": 1.4043,
+      "step": 13284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7371544362917435,
+      "learning_rate": 1.7697000314166557e-05,
+      "loss": 1.4473,
+      "step": 13285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.208662639291642,
+      "learning_rate": 1.7693011777248276e-05,
+      "loss": 1.1973,
+      "step": 13286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8489679107412047,
+      "learning_rate": 1.7689023443680492e-05,
+      "loss": 1.0605,
+      "step": 13287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3980246322104852,
+      "learning_rate": 1.7685035313574216e-05,
+      "loss": 1.252,
+      "step": 13288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4003599334070724,
+      "learning_rate": 1.7681047387040427e-05,
+      "loss": 0.7188,
+      "step": 13289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.44001380208118,
+      "learning_rate": 1.7677059664190116e-05,
+      "loss": 1.6191,
+      "step": 13290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.898541657771709,
+      "learning_rate": 1.7673072145134242e-05,
+      "loss": 1.1484,
+      "step": 13291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.289803307210108,
+      "learning_rate": 1.7669084829983796e-05,
+      "loss": 1.2026,
+      "step": 13292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.7082406544073825,
+      "learning_rate": 1.766509771884972e-05,
+      "loss": 1.1074,
+      "step": 13293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.94424311034532,
+      "learning_rate": 1.7661110811843e-05,
+      "loss": 1.0762,
+      "step": 13294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9151723329974082,
+      "learning_rate": 1.765712410907458e-05,
+      "loss": 1.0771,
+      "step": 13295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.875823690758804,
+      "learning_rate": 1.7653137610655406e-05,
+      "loss": 0.8784,
+      "step": 13296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.337751723826424,
+      "learning_rate": 1.7649151316696423e-05,
+      "loss": 1.3242,
+      "step": 13297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0597545962777235,
+      "learning_rate": 1.764516522730856e-05,
+      "loss": 0.6289,
+      "step": 13298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.204669115003562,
+      "learning_rate": 1.764117934260276e-05,
+      "loss": 1.207,
+      "step": 13299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.400910038074166,
+      "learning_rate": 1.7637193662689943e-05,
+      "loss": 0.8984,
+      "step": 13300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7213057858974405,
+      "learning_rate": 1.7633208187681028e-05,
+      "loss": 0.7744,
+      "step": 13301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0677683042450097,
+      "learning_rate": 1.762922291768692e-05,
+      "loss": 0.7017,
+      "step": 13302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.113103130551664,
+      "learning_rate": 1.7625237852818555e-05,
+      "loss": 0.9893,
+      "step": 13303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.846246690705791,
+      "learning_rate": 1.7621252993186804e-05,
+      "loss": 1.1133,
+      "step": 13304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7616714680553303,
+      "learning_rate": 1.761726833890259e-05,
+      "loss": 1.1172,
+      "step": 13305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.400897923026298,
+      "learning_rate": 1.7613283890076786e-05,
+      "loss": 1.4375,
+      "step": 13306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.39217471638272,
+      "learning_rate": 1.7609299646820286e-05,
+      "loss": 1.1533,
+      "step": 13307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.438675425849369,
+      "learning_rate": 1.7605315609243965e-05,
+      "loss": 0.9263,
+      "step": 13308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7761988687766195,
+      "learning_rate": 1.7601331777458706e-05,
+      "loss": 0.5864,
+      "step": 13309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5526307018566228,
+      "learning_rate": 1.759734815157537e-05,
+      "loss": 1.1719,
+      "step": 13310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.188617638769205,
+      "learning_rate": 1.7593364731704812e-05,
+      "loss": 1.2305,
+      "step": 13311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.59236264180378,
+      "learning_rate": 1.7589381517957902e-05,
+      "loss": 1.5742,
+      "step": 13312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.693863878561665,
+      "learning_rate": 1.7585398510445497e-05,
+      "loss": 0.9639,
+      "step": 13313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2110208096799613,
+      "learning_rate": 1.7581415709278426e-05,
+      "loss": 1.0723,
+      "step": 13314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.032361020454072,
+      "learning_rate": 1.757743311456754e-05,
+      "loss": 1.2578,
+      "step": 13315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6967570604486646,
+      "learning_rate": 1.7573450726423662e-05,
+      "loss": 0.6089,
+      "step": 13316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.812249405611165,
+      "learning_rate": 1.7569468544957635e-05,
+      "loss": 1.4395,
+      "step": 13317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4432277384791568,
+      "learning_rate": 1.7565486570280268e-05,
+      "loss": 0.7773,
+      "step": 13318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1563077109320106,
+      "learning_rate": 1.7561504802502388e-05,
+      "loss": 0.959,
+      "step": 13319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.091973071982045,
+      "learning_rate": 1.7557523241734793e-05,
+      "loss": 0.7812,
+      "step": 13320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.151070300936314,
+      "learning_rate": 1.7553541888088295e-05,
+      "loss": 0.7803,
+      "step": 13321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9677620206194963,
+      "learning_rate": 1.7549560741673703e-05,
+      "loss": 0.6636,
+      "step": 13322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3899892323143663,
+      "learning_rate": 1.7545579802601803e-05,
+      "loss": 0.9297,
+      "step": 13323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.01707973768574,
+      "learning_rate": 1.754159907098338e-05,
+      "loss": 1.1152,
+      "step": 13324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3088292035704514,
+      "learning_rate": 1.7537618546929217e-05,
+      "loss": 1.0459,
+      "step": 13325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.472860976064234,
+      "learning_rate": 1.7533638230550098e-05,
+      "loss": 0.7744,
+      "step": 13326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.785561743590847,
+      "learning_rate": 1.7529658121956777e-05,
+      "loss": 1.6836,
+      "step": 13327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.12353252707839,
+      "learning_rate": 1.752567822126004e-05,
+      "loss": 0.7395,
+      "step": 13328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.812288341975445,
+      "learning_rate": 1.752169852857063e-05,
+      "loss": 1.2461,
+      "step": 13329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.238800361445979,
+      "learning_rate": 1.7517719043999298e-05,
+      "loss": 1.1084,
+      "step": 13330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2030882298289645,
+      "learning_rate": 1.7513739767656807e-05,
+      "loss": 1.3496,
+      "step": 13331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9891624198679927,
+      "learning_rate": 1.7509760699653887e-05,
+      "loss": 0.5854,
+      "step": 13332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2216753949134356,
+      "learning_rate": 1.7505781840101283e-05,
+      "loss": 1.0444,
+      "step": 13333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.345815430653056,
+      "learning_rate": 1.7501803189109718e-05,
+      "loss": 0.7705,
+      "step": 13334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3507633990548054,
+      "learning_rate": 1.7497824746789914e-05,
+      "loss": 0.8794,
+      "step": 13335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.115354751162698,
+      "learning_rate": 1.74938465132526e-05,
+      "loss": 0.9014,
+      "step": 13336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6283886245430295,
+      "learning_rate": 1.7489868488608478e-05,
+      "loss": 1.0586,
+      "step": 13337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.707302063445828,
+      "learning_rate": 1.748589067296826e-05,
+      "loss": 1.0107,
+      "step": 13338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.225420511908129,
+      "learning_rate": 1.7481913066442638e-05,
+      "loss": 0.8984,
+      "step": 13339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4729643643190578,
+      "learning_rate": 1.7477935669142316e-05,
+      "loss": 1.1328,
+      "step": 13340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.780870011133381,
+      "learning_rate": 1.747395848117799e-05,
+      "loss": 0.9932,
+      "step": 13341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7554876254348275,
+      "learning_rate": 1.746998150266033e-05,
+      "loss": 1.1387,
+      "step": 13342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3008000060581533,
+      "learning_rate": 1.7466004733700026e-05,
+      "loss": 0.9629,
+      "step": 13343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8378506180325735,
+      "learning_rate": 1.746202817440774e-05,
+      "loss": 0.9492,
+      "step": 13344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.924265816646334,
+      "learning_rate": 1.7458051824894143e-05,
+      "loss": 0.6733,
+      "step": 13345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.58276079868049,
+      "learning_rate": 1.745407568526989e-05,
+      "loss": 1.3633,
+      "step": 13346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.755384775250156,
+      "learning_rate": 1.745009975564564e-05,
+      "loss": 1.2168,
+      "step": 13347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7569862403125014,
+      "learning_rate": 1.7446124036132035e-05,
+      "loss": 1.4746,
+      "step": 13348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.68682773077328,
+      "learning_rate": 1.7442148526839726e-05,
+      "loss": 1.3359,
+      "step": 13349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5509631917387001,
+      "learning_rate": 1.743817322787935e-05,
+      "loss": 0.5918,
+      "step": 13350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.259356199832709,
+      "learning_rate": 1.743419813936153e-05,
+      "loss": 1.1504,
+      "step": 13351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.454820180860775,
+      "learning_rate": 1.7430223261396893e-05,
+      "loss": 1.2783,
+      "step": 13352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8127537911411888,
+      "learning_rate": 1.7426248594096072e-05,
+      "loss": 0.6274,
+      "step": 13353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5723267984610847,
+      "learning_rate": 1.742227413756966e-05,
+      "loss": 0.9482,
+      "step": 13354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.947118606358767,
+      "learning_rate": 1.741829989192828e-05,
+      "loss": 1.0342,
+      "step": 13355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.159965851655105,
+      "learning_rate": 1.7414325857282526e-05,
+      "loss": 1.2769,
+      "step": 13356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.337574872944414,
+      "learning_rate": 1.7410352033742988e-05,
+      "loss": 1.0234,
+      "step": 13357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.065630415593376,
+      "learning_rate": 1.7406378421420273e-05,
+      "loss": 1.1279,
+      "step": 13358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.175014499514995,
+      "learning_rate": 1.740240502042495e-05,
+      "loss": 1.3184,
+      "step": 13359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6770665110097878,
+      "learning_rate": 1.7398431830867605e-05,
+      "loss": 0.918,
+      "step": 13360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.526036492605584,
+      "learning_rate": 1.7394458852858805e-05,
+      "loss": 1.2246,
+      "step": 13361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.72184823493769,
+      "learning_rate": 1.7390486086509122e-05,
+      "loss": 0.6797,
+      "step": 13362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5495773675778857,
+      "learning_rate": 1.738651353192911e-05,
+      "loss": 0.9287,
+      "step": 13363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3781690034542287,
+      "learning_rate": 1.7382541189229333e-05,
+      "loss": 0.9902,
+      "step": 13364
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4060206904788064,
+      "learning_rate": 1.737856905852033e-05,
+      "loss": 0.5962,
+      "step": 13365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.105720864864404,
+      "learning_rate": 1.7374597139912648e-05,
+      "loss": 1.3359,
+      "step": 13366
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.011492848257281,
+      "learning_rate": 1.7370625433516817e-05,
+      "loss": 2.0977,
+      "step": 13367
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.548018497190302,
+      "learning_rate": 1.7366653939443377e-05,
+      "loss": 1.2988,
+      "step": 13368
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4058987249620816,
+      "learning_rate": 1.7362682657802852e-05,
+      "loss": 1.0977,
+      "step": 13369
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5579160302333244,
+      "learning_rate": 1.735871158870576e-05,
+      "loss": 1.2812,
+      "step": 13370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9949101012311834,
+      "learning_rate": 1.735474073226262e-05,
+      "loss": 1.1953,
+      "step": 13371
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9011649801592858,
+      "learning_rate": 1.7350770088583923e-05,
+      "loss": 0.6682,
+      "step": 13372
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8253965757737953,
+      "learning_rate": 1.734679965778018e-05,
+      "loss": 0.7422,
+      "step": 13373
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.0129577579906295,
+      "learning_rate": 1.7342829439961898e-05,
+      "loss": 0.7803,
+      "step": 13374
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8854962597314655,
+      "learning_rate": 1.7338859435239547e-05,
+      "loss": 1.5703,
+      "step": 13375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.399594304445594,
+      "learning_rate": 1.7334889643723612e-05,
+      "loss": 1.2109,
+      "step": 13376
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9278147022706549,
+      "learning_rate": 1.733092006552459e-05,
+      "loss": 1.1914,
+      "step": 13377
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.861877168293664,
+      "learning_rate": 1.7326950700752938e-05,
+      "loss": 1.1055,
+      "step": 13378
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.868654296874335,
+      "learning_rate": 1.7322981549519124e-05,
+      "loss": 0.877,
+      "step": 13379
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8755300836085738,
+      "learning_rate": 1.7319012611933604e-05,
+      "loss": 0.7803,
+      "step": 13380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1851170230490964,
+      "learning_rate": 1.7315043888106846e-05,
+      "loss": 0.7002,
+      "step": 13381
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.196249765169221,
+      "learning_rate": 1.7311075378149276e-05,
+      "loss": 1.1406,
+      "step": 13382
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.145225426448125,
+      "learning_rate": 1.730710708217136e-05,
+      "loss": 0.7617,
+      "step": 13383
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.422594663825633,
+      "learning_rate": 1.7303139000283513e-05,
+      "loss": 1.0947,
+      "step": 13384
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.751688438882381,
+      "learning_rate": 1.729917113259617e-05,
+      "loss": 1.1387,
+      "step": 13385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8163246428667548,
+      "learning_rate": 1.729520347921977e-05,
+      "loss": 1.6035,
+      "step": 13386
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9787290775229542,
+      "learning_rate": 1.7291236040264717e-05,
+      "loss": 1.0586,
+      "step": 13387
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.512597826481687,
+      "learning_rate": 1.728726881584143e-05,
+      "loss": 0.9195,
+      "step": 13388
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.015367733070508,
+      "learning_rate": 1.7283301806060304e-05,
+      "loss": 0.7793,
+      "step": 13389
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.891048507239474,
+      "learning_rate": 1.7279335011031755e-05,
+      "loss": 1.3262,
+      "step": 13390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6633899544564463,
+      "learning_rate": 1.7275368430866168e-05,
+      "loss": 1.3145,
+      "step": 13391
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3450701237785143,
+      "learning_rate": 1.7271402065673928e-05,
+      "loss": 1.2607,
+      "step": 13392
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4810026300148054,
+      "learning_rate": 1.7267435915565427e-05,
+      "loss": 1.2969,
+      "step": 13393
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2391575638787335,
+      "learning_rate": 1.7263469980651033e-05,
+      "loss": 0.7998,
+      "step": 13394
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.054510188701559,
+      "learning_rate": 1.7259504261041115e-05,
+      "loss": 0.8815,
+      "step": 13395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9709525520904,
+      "learning_rate": 1.7255538756846054e-05,
+      "loss": 0.8086,
+      "step": 13396
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3112556615072077,
+      "learning_rate": 1.725157346817619e-05,
+      "loss": 0.8311,
+      "step": 13397
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.474331555394311,
+      "learning_rate": 1.7247608395141886e-05,
+      "loss": 1.709,
+      "step": 13398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0244647036749694,
+      "learning_rate": 1.7243643537853483e-05,
+      "loss": 1.2129,
+      "step": 13399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.045816090783308,
+      "learning_rate": 1.7239678896421324e-05,
+      "loss": 1.1182,
+      "step": 13400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.858841309484054,
+      "learning_rate": 1.723571447095574e-05,
+      "loss": 1.4062,
+      "step": 13401
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.010098558929797,
+      "learning_rate": 1.7231750261567064e-05,
+      "loss": 1.6875,
+      "step": 13402
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.60479617838919,
+      "learning_rate": 1.7227786268365608e-05,
+      "loss": 0.9482,
+      "step": 13403
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.467348594310489,
+      "learning_rate": 1.7223822491461695e-05,
+      "loss": 0.7372,
+      "step": 13404
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4694944390498246,
+      "learning_rate": 1.7219858930965648e-05,
+      "loss": 1.5977,
+      "step": 13405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8287177544819637,
+      "learning_rate": 1.7215895586987753e-05,
+      "loss": 1.0254,
+      "step": 13406
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.532931847224595,
+      "learning_rate": 1.721193245963832e-05,
+      "loss": 1.165,
+      "step": 13407
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6026736715414795,
+      "learning_rate": 1.720796954902763e-05,
+      "loss": 1.4512,
+      "step": 13408
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.399849739934039,
+      "learning_rate": 1.7204006855265982e-05,
+      "loss": 1.0029,
+      "step": 13409
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.400340962032017,
+      "learning_rate": 1.7200044378463643e-05,
+      "loss": 1.2705,
+      "step": 13410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7653202384661983,
+      "learning_rate": 1.71960821187309e-05,
+      "loss": 1.3066,
+      "step": 13411
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2002651939177427,
+      "learning_rate": 1.719212007617801e-05,
+      "loss": 0.9365,
+      "step": 13412
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4784674312875046,
+      "learning_rate": 1.7188158250915233e-05,
+      "loss": 1.5879,
+      "step": 13413
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8737330288719676,
+      "learning_rate": 1.7184196643052834e-05,
+      "loss": 1.0944,
+      "step": 13414
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3331336094891313,
+      "learning_rate": 1.718023525270107e-05,
+      "loss": 1.1211,
+      "step": 13415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.123606393145065,
+      "learning_rate": 1.717627407997017e-05,
+      "loss": 0.751,
+      "step": 13416
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7542233548315505,
+      "learning_rate": 1.717231312497038e-05,
+      "loss": 0.6416,
+      "step": 13417
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4433261981910235,
+      "learning_rate": 1.7168352387811926e-05,
+      "loss": 0.9316,
+      "step": 13418
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.846710875220782,
+      "learning_rate": 1.716439186860504e-05,
+      "loss": 1.1836,
+      "step": 13419
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.290980416401054,
+      "learning_rate": 1.7160431567459932e-05,
+      "loss": 1.082,
+      "step": 13420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.042941013743539,
+      "learning_rate": 1.715647148448683e-05,
+      "loss": 0.8779,
+      "step": 13421
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5827710501547747,
+      "learning_rate": 1.7152511619795924e-05,
+      "loss": 0.7441,
+      "step": 13422
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9658353420593686,
+      "learning_rate": 1.714855197349743e-05,
+      "loss": 0.7993,
+      "step": 13423
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.3208440141884,
+      "learning_rate": 1.7144592545701542e-05,
+      "loss": 1.5645,
+      "step": 13424
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.336052320841558,
+      "learning_rate": 1.714063333651844e-05,
+      "loss": 1.0586,
+      "step": 13425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.481082997209588,
+      "learning_rate": 1.713667434605832e-05,
+      "loss": 1.0537,
+      "step": 13426
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6326954982579833,
+      "learning_rate": 1.7132715574431345e-05,
+      "loss": 0.8672,
+      "step": 13427
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.313322576918562,
+      "learning_rate": 1.71287570217477e-05,
+      "loss": 0.9551,
+      "step": 13428
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.532143517547343,
+      "learning_rate": 1.7124798688117532e-05,
+      "loss": 0.483,
+      "step": 13429
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9753814908269638,
+      "learning_rate": 1.712084057365102e-05,
+      "loss": 1.0215,
+      "step": 13430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.05666671986973,
+      "learning_rate": 1.7116882678458303e-05,
+      "loss": 0.9824,
+      "step": 13431
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.048385707403864,
+      "learning_rate": 1.7112925002649528e-05,
+      "loss": 0.8506,
+      "step": 13432
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7511088956984033,
+      "learning_rate": 1.710896754633484e-05,
+      "loss": 1.7266,
+      "step": 13433
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1267728735719316,
+      "learning_rate": 1.710501030962438e-05,
+      "loss": 0.9424,
+      "step": 13434
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.070235456076775,
+      "learning_rate": 1.710105329262826e-05,
+      "loss": 1.4746,
+      "step": 13435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0380372725682525,
+      "learning_rate": 1.709709649545662e-05,
+      "loss": 1.082,
+      "step": 13436
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.782268340925584,
+      "learning_rate": 1.7093139918219565e-05,
+      "loss": 0.8506,
+      "step": 13437
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5990334438155145,
+      "learning_rate": 1.708918356102721e-05,
+      "loss": 0.8232,
+      "step": 13438
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.008518911100306,
+      "learning_rate": 1.7085227423989647e-05,
+      "loss": 1.2266,
+      "step": 13439
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.629825335022197,
+      "learning_rate": 1.7081271507216992e-05,
+      "loss": 0.542,
+      "step": 13440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.169977141053355,
+      "learning_rate": 1.7077315810819315e-05,
+      "loss": 0.8086,
+      "step": 13441
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.700089058355313,
+      "learning_rate": 1.7073360334906723e-05,
+      "loss": 1.4043,
+      "step": 13442
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.711716070149119,
+      "learning_rate": 1.7069405079589286e-05,
+      "loss": 0.6787,
+      "step": 13443
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0856660244764784,
+      "learning_rate": 1.706545004497707e-05,
+      "loss": 0.5957,
+      "step": 13444
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.539277052984322,
+      "learning_rate": 1.7061495231180162e-05,
+      "loss": 1.1104,
+      "step": 13445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.427302288549904,
+      "learning_rate": 1.70575406383086e-05,
+      "loss": 0.9141,
+      "step": 13446
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8191078358374335,
+      "learning_rate": 1.7053586266472456e-05,
+      "loss": 0.3188,
+      "step": 13447
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.334651945397106,
+      "learning_rate": 1.7049632115781765e-05,
+      "loss": 2.1411,
+      "step": 13448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2884576929376963,
+      "learning_rate": 1.704567818634658e-05,
+      "loss": 0.688,
+      "step": 13449
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7540491806277592,
+      "learning_rate": 1.7041724478276915e-05,
+      "loss": 0.7139,
+      "step": 13450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9696392133035188,
+      "learning_rate": 1.703777099168284e-05,
+      "loss": 0.9756,
+      "step": 13451
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8888716961819962,
+      "learning_rate": 1.7033817726674345e-05,
+      "loss": 1.2217,
+      "step": 13452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5832798142170468,
+      "learning_rate": 1.7029864683361468e-05,
+      "loss": 1.0195,
+      "step": 13453
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.174229320645866,
+      "learning_rate": 1.7025911861854207e-05,
+      "loss": 1.3652,
+      "step": 13454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.58265082114044,
+      "learning_rate": 1.7021959262262578e-05,
+      "loss": 1.2725,
+      "step": 13455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4011850668514025,
+      "learning_rate": 1.701800688469657e-05,
+      "loss": 0.7363,
+      "step": 13456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.773080266143738,
+      "learning_rate": 1.7014054729266183e-05,
+      "loss": 0.6763,
+      "step": 13457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2680393766741362,
+      "learning_rate": 1.7010102796081402e-05,
+      "loss": 0.6509,
+      "step": 13458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.482842698582768,
+      "learning_rate": 1.70061510852522e-05,
+      "loss": 1.084,
+      "step": 13459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.209345483650189,
+      "learning_rate": 1.700219959688857e-05,
+      "loss": 0.9414,
+      "step": 13460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.369529321549291,
+      "learning_rate": 1.6998248331100457e-05,
+      "loss": 1.4238,
+      "step": 13461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.578851308529184,
+      "learning_rate": 1.6994297287997852e-05,
+      "loss": 1.1982,
+      "step": 13462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6229937016145723,
+      "learning_rate": 1.6990346467690683e-05,
+      "loss": 0.7739,
+      "step": 13463
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5559504475692965,
+      "learning_rate": 1.6986395870288915e-05,
+      "loss": 0.999,
+      "step": 13464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.705533228300565,
+      "learning_rate": 1.698244549590248e-05,
+      "loss": 1.0586,
+      "step": 13465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.000450100512265,
+      "learning_rate": 1.6978495344641338e-05,
+      "loss": 1.1367,
+      "step": 13466
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.325747721870924,
+      "learning_rate": 1.6974545416615396e-05,
+      "loss": 1.5156,
+      "step": 13467
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0745461990313165,
+      "learning_rate": 1.697059571193459e-05,
+      "loss": 1.1064,
+      "step": 13468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.741197937036927,
+      "learning_rate": 1.696664623070882e-05,
+      "loss": 0.9116,
+      "step": 13469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.602573965672379,
+      "learning_rate": 1.6962696973048033e-05,
+      "loss": 1.2539,
+      "step": 13470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7162396217693654,
+      "learning_rate": 1.6958747939062114e-05,
+      "loss": 0.7289,
+      "step": 13471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4805047217523954,
+      "learning_rate": 1.695479912886097e-05,
+      "loss": 0.8169,
+      "step": 13472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.692747916824881,
+      "learning_rate": 1.6950850542554483e-05,
+      "loss": 0.5825,
+      "step": 13473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.350436788523126,
+      "learning_rate": 1.6946902180252556e-05,
+      "loss": 1.0371,
+      "step": 13474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6644906939277608,
+      "learning_rate": 1.6942954042065057e-05,
+      "loss": 0.8972,
+      "step": 13475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.299273425899656,
+      "learning_rate": 1.6939006128101876e-05,
+      "loss": 1.4648,
+      "step": 13476
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3960164590931825,
+      "learning_rate": 1.693505843847286e-05,
+      "loss": 1.2151,
+      "step": 13477
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.29436050361515,
+      "learning_rate": 1.6931110973287886e-05,
+      "loss": 2.0605,
+      "step": 13478
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9778192179372382,
+      "learning_rate": 1.6927163732656816e-05,
+      "loss": 0.9951,
+      "step": 13479
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.334937554612896,
+      "learning_rate": 1.692321671668949e-05,
+      "loss": 0.9961,
+      "step": 13480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.276881832347451,
+      "learning_rate": 1.6919269925495755e-05,
+      "loss": 0.8457,
+      "step": 13481
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.506315965689327,
+      "learning_rate": 1.6915323359185446e-05,
+      "loss": 1.4736,
+      "step": 13482
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7465812934026785,
+      "learning_rate": 1.6911377017868406e-05,
+      "loss": 1.3887,
+      "step": 13483
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6464813340997275,
+      "learning_rate": 1.690743090165444e-05,
+      "loss": 1.0967,
+      "step": 13484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7927531979485942,
+      "learning_rate": 1.6903485010653385e-05,
+      "loss": 0.752,
+      "step": 13485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.651751912200513,
+      "learning_rate": 1.6899539344975042e-05,
+      "loss": 0.9473,
+      "step": 13486
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8844203032038633,
+      "learning_rate": 1.6895593904729215e-05,
+      "loss": 1.0791,
+      "step": 13487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.107036509381761,
+      "learning_rate": 1.6891648690025718e-05,
+      "loss": 1.457,
+      "step": 13488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4271140208314903,
+      "learning_rate": 1.6887703700974336e-05,
+      "loss": 1.2949,
+      "step": 13489
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.651346415232856,
+      "learning_rate": 1.6883758937684857e-05,
+      "loss": 1.2285,
+      "step": 13490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.683729155478558,
+      "learning_rate": 1.6879814400267065e-05,
+      "loss": 1.3145,
+      "step": 13491
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1178315369030587,
+      "learning_rate": 1.687587008883073e-05,
+      "loss": 0.9844,
+      "step": 13492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.155024083261827,
+      "learning_rate": 1.6871926003485627e-05,
+      "loss": 1.082,
+      "step": 13493
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7674034635836224,
+      "learning_rate": 1.6867982144341508e-05,
+      "loss": 0.9453,
+      "step": 13494
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.499361001088043,
+      "learning_rate": 1.6864038511508144e-05,
+      "loss": 1.1631,
+      "step": 13495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.949647339253348,
+      "learning_rate": 1.6860095105095265e-05,
+      "loss": 0.853,
+      "step": 13496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3585773660773692,
+      "learning_rate": 1.6856151925212627e-05,
+      "loss": 0.7317,
+      "step": 13497
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.71082395816014,
+      "learning_rate": 1.6852208971969975e-05,
+      "loss": 1.0,
+      "step": 13498
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.169511337161305,
+      "learning_rate": 1.6848266245477022e-05,
+      "loss": 1.1855,
+      "step": 13499
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.101899300853624,
+      "learning_rate": 1.684432374584351e-05,
+      "loss": 1.0059,
+      "step": 13500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.18691711680541,
+      "learning_rate": 1.6840381473179137e-05,
+      "loss": 1.0586,
+      "step": 13501
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6919974363324326,
+      "learning_rate": 1.6836439427593633e-05,
+      "loss": 1.2402,
+      "step": 13502
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.688756954816094,
+      "learning_rate": 1.6832497609196694e-05,
+      "loss": 0.8984,
+      "step": 13503
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.88244627532856,
+      "learning_rate": 1.6828556018098024e-05,
+      "loss": 1.1465,
+      "step": 13504
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7942296998320135,
+      "learning_rate": 1.6824614654407307e-05,
+      "loss": 1.0576,
+      "step": 13505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9843837877713386,
+      "learning_rate": 1.6820673518234238e-05,
+      "loss": 1.3047,
+      "step": 13506
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8885662959283067,
+      "learning_rate": 1.6816732609688497e-05,
+      "loss": 0.9355,
+      "step": 13507
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.247274429674624,
+      "learning_rate": 1.6812791928879757e-05,
+      "loss": 1.1797,
+      "step": 13508
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8376870767248894,
+      "learning_rate": 1.680885147591768e-05,
+      "loss": 0.7026,
+      "step": 13509
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5461306742226104,
+      "learning_rate": 1.6804911250911937e-05,
+      "loss": 0.468,
+      "step": 13510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.038593170743866,
+      "learning_rate": 1.6800971253972175e-05,
+      "loss": 1.2422,
+      "step": 13511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7670484203105694,
+      "learning_rate": 1.679703148520805e-05,
+      "loss": 1.5469,
+      "step": 13512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.633066046304046,
+      "learning_rate": 1.6793091944729194e-05,
+      "loss": 1.1445,
+      "step": 13513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.366090387809182,
+      "learning_rate": 1.6789152632645256e-05,
+      "loss": 1.0068,
+      "step": 13514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.491823187041313,
+      "learning_rate": 1.6785213549065844e-05,
+      "loss": 1.1934,
+      "step": 13515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0141870855207356,
+      "learning_rate": 1.67812746941006e-05,
+      "loss": 0.5742,
+      "step": 13516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3684208744450554,
+      "learning_rate": 1.677733606785914e-05,
+      "loss": 1.0762,
+      "step": 13517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4163580189058473,
+      "learning_rate": 1.6773397670451063e-05,
+      "loss": 0.5986,
+      "step": 13518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7093402195676193,
+      "learning_rate": 1.676945950198599e-05,
+      "loss": 1.0537,
+      "step": 13519
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4725326670571866,
+      "learning_rate": 1.6765521562573498e-05,
+      "loss": 0.5581,
+      "step": 13520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3266963847584208,
+      "learning_rate": 1.67615838523232e-05,
+      "loss": 1.0811,
+      "step": 13521
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8922830672507758,
+      "learning_rate": 1.675764637134466e-05,
+      "loss": 0.7158,
+      "step": 13522
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.900465921265893,
+      "learning_rate": 1.6753709119747473e-05,
+      "loss": 1.2188,
+      "step": 13523
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.374316504734942,
+      "learning_rate": 1.674977209764119e-05,
+      "loss": 1.2842,
+      "step": 13524
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.690254734209677,
+      "learning_rate": 1.67458353051354e-05,
+      "loss": 0.6499,
+      "step": 13525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.010826479627686,
+      "learning_rate": 1.6741898742339656e-05,
+      "loss": 0.8369,
+      "step": 13526
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.969538452646425,
+      "learning_rate": 1.6737962409363506e-05,
+      "loss": 1.4121,
+      "step": 13527
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.006541590517293,
+      "learning_rate": 1.6734026306316503e-05,
+      "loss": 0.6489,
+      "step": 13528
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.241568409342006,
+      "learning_rate": 1.6730090433308178e-05,
+      "loss": 1.3164,
+      "step": 13529
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4065362338951077,
+      "learning_rate": 1.6726154790448067e-05,
+      "loss": 0.7861,
+      "step": 13530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8050621271404799,
+      "learning_rate": 1.6722219377845703e-05,
+      "loss": 1.4316,
+      "step": 13531
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.023768049598619,
+      "learning_rate": 1.6718284195610606e-05,
+      "loss": 1.0947,
+      "step": 13532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.397806183503812,
+      "learning_rate": 1.671434924385228e-05,
+      "loss": 0.9009,
+      "step": 13533
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0221000581504662,
+      "learning_rate": 1.6710414522680247e-05,
+      "loss": 0.2362,
+      "step": 13534
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7733103164574975,
+      "learning_rate": 1.6706480032204e-05,
+      "loss": 0.4829,
+      "step": 13535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6358532720825356,
+      "learning_rate": 1.670254577253305e-05,
+      "loss": 1.2988,
+      "step": 13536
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9344461388802054,
+      "learning_rate": 1.669861174377686e-05,
+      "loss": 1.2861,
+      "step": 13537
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.499408169895758,
+      "learning_rate": 1.6694677946044934e-05,
+      "loss": 1.4922,
+      "step": 13538
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.570188934151277,
+      "learning_rate": 1.6690744379446735e-05,
+      "loss": 1.832,
+      "step": 13539
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.12909102731661,
+      "learning_rate": 1.6686811044091742e-05,
+      "loss": 0.3105,
+      "step": 13540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.11551719370169,
+      "learning_rate": 1.6682877940089406e-05,
+      "loss": 1.5371,
+      "step": 13541
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4993784131739614,
+      "learning_rate": 1.6678945067549197e-05,
+      "loss": 1.6211,
+      "step": 13542
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5149344687143422,
+      "learning_rate": 1.667501242658055e-05,
+      "loss": 0.7422,
+      "step": 13543
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0829773431045036,
+      "learning_rate": 1.6671080017292922e-05,
+      "loss": 0.6479,
+      "step": 13544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.399225062191509,
+      "learning_rate": 1.6667147839795756e-05,
+      "loss": 1.0703,
+      "step": 13545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.107219159547093,
+      "learning_rate": 1.6663215894198463e-05,
+      "loss": 1.4629,
+      "step": 13546
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7771026893011768,
+      "learning_rate": 1.6659284180610485e-05,
+      "loss": 0.7563,
+      "step": 13547
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4077859905548618,
+      "learning_rate": 1.6655352699141224e-05,
+      "loss": 1.0918,
+      "step": 13548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7344192065343984,
+      "learning_rate": 1.6651421449900105e-05,
+      "loss": 0.7637,
+      "step": 13549
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.214767876986676,
+      "learning_rate": 1.6647490432996533e-05,
+      "loss": 1.2285,
+      "step": 13550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.655246498473293,
+      "learning_rate": 1.6643559648539896e-05,
+      "loss": 1.0977,
+      "step": 13551
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8706668873792895,
+      "learning_rate": 1.6639629096639587e-05,
+      "loss": 1.4766,
+      "step": 13552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5601267175306877,
+      "learning_rate": 1.6635698777405008e-05,
+      "loss": 0.9004,
+      "step": 13553
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.775995670436154,
+      "learning_rate": 1.663176869094552e-05,
+      "loss": 1.0547,
+      "step": 13554
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9917525729823944,
+      "learning_rate": 1.6627838837370508e-05,
+      "loss": 0.5417,
+      "step": 13555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.479675358320099,
+      "learning_rate": 1.6623909216789328e-05,
+      "loss": 0.9561,
+      "step": 13556
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.922332128581282,
+      "learning_rate": 1.661997982931135e-05,
+      "loss": 1.5977,
+      "step": 13557
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.364209905319277,
+      "learning_rate": 1.6616050675045914e-05,
+      "loss": 1.1074,
+      "step": 13558
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7340902125691295,
+      "learning_rate": 1.6612121754102385e-05,
+      "loss": 1.0356,
+      "step": 13559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.310891906825003,
+      "learning_rate": 1.6608193066590084e-05,
+      "loss": 0.9883,
+      "step": 13560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.852111992401502,
+      "learning_rate": 1.660426461261835e-05,
+      "loss": 0.7405,
+      "step": 13561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1300277695614165,
+      "learning_rate": 1.6600336392296523e-05,
+      "loss": 0.9375,
+      "step": 13562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.12621749585731,
+      "learning_rate": 1.659640840573391e-05,
+      "loss": 1.0132,
+      "step": 13563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7926453944621246,
+      "learning_rate": 1.6592480653039833e-05,
+      "loss": 1.4927,
+      "step": 13564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4301085275975023,
+      "learning_rate": 1.6588553134323593e-05,
+      "loss": 1.3652,
+      "step": 13565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.62921266114221,
+      "learning_rate": 1.6584625849694503e-05,
+      "loss": 0.7529,
+      "step": 13566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5703655573693682,
+      "learning_rate": 1.658069879926184e-05,
+      "loss": 0.9062,
+      "step": 13567
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2300527689966505,
+      "learning_rate": 1.657677198313491e-05,
+      "loss": 0.5977,
+      "step": 13568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9957270873373338,
+      "learning_rate": 1.6572845401422983e-05,
+      "loss": 1.2324,
+      "step": 13569
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.0428092888238965,
+      "learning_rate": 1.6568919054235335e-05,
+      "loss": 0.5469,
+      "step": 13570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.646461942795843,
+      "learning_rate": 1.6564992941681235e-05,
+      "loss": 0.8252,
+      "step": 13571
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.088985434821687,
+      "learning_rate": 1.6561067063869953e-05,
+      "loss": 0.8828,
+      "step": 13572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.268393357674189,
+      "learning_rate": 1.655714142091074e-05,
+      "loss": 0.8955,
+      "step": 13573
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2887048232347755,
+      "learning_rate": 1.655321601291284e-05,
+      "loss": 1.0625,
+      "step": 13574
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.607812141144085,
+      "learning_rate": 1.6549290839985505e-05,
+      "loss": 0.8553,
+      "step": 13575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.429022400521475,
+      "learning_rate": 1.6545365902237964e-05,
+      "loss": 0.6519,
+      "step": 13576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5597480325081214,
+      "learning_rate": 1.6541441199779445e-05,
+      "loss": 0.9629,
+      "step": 13577
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8204711700356375,
+      "learning_rate": 1.6537516732719182e-05,
+      "loss": 1.002,
+      "step": 13578
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.340734041778262,
+      "learning_rate": 1.6533592501166374e-05,
+      "loss": 1.3223,
+      "step": 13579
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.09926914777196,
+      "learning_rate": 1.6529668505230238e-05,
+      "loss": 2.4375,
+      "step": 13580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.538847553251499,
+      "learning_rate": 1.652574474501999e-05,
+      "loss": 0.6346,
+      "step": 13581
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4042354989466956,
+      "learning_rate": 1.6521821220644807e-05,
+      "loss": 0.1425,
+      "step": 13582
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.750974489646638,
+      "learning_rate": 1.6517897932213892e-05,
+      "loss": 1.1123,
+      "step": 13583
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.266556962481279,
+      "learning_rate": 1.6513974879836426e-05,
+      "loss": 0.9819,
+      "step": 13584
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.239594680558499,
+      "learning_rate": 1.6510052063621585e-05,
+      "loss": 1.2002,
+      "step": 13585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7245931159926324,
+      "learning_rate": 1.6506129483678534e-05,
+      "loss": 1.2773,
+      "step": 13586
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0798113871301833,
+      "learning_rate": 1.6502207140116446e-05,
+      "loss": 0.6746,
+      "step": 13587
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.681060436775426,
+      "learning_rate": 1.649828503304447e-05,
+      "loss": 1.2949,
+      "step": 13588
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7994832924509327,
+      "learning_rate": 1.6494363162571753e-05,
+      "loss": 0.8799,
+      "step": 13589
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.432475468466443,
+      "learning_rate": 1.6490441528807456e-05,
+      "loss": 1.4873,
+      "step": 13590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.496613181672493,
+      "learning_rate": 1.6486520131860712e-05,
+      "loss": 0.6345,
+      "step": 13591
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.831929887733467,
+      "learning_rate": 1.6482598971840638e-05,
+      "loss": 0.7202,
+      "step": 13592
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9740640702019165,
+      "learning_rate": 1.647867804885637e-05,
+      "loss": 1.3926,
+      "step": 13593
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3930076278929073,
+      "learning_rate": 1.6474757363017023e-05,
+      "loss": 0.8359,
+      "step": 13594
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.606338470201528,
+      "learning_rate": 1.6470836914431715e-05,
+      "loss": 0.7061,
+      "step": 13595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2301860717937974,
+      "learning_rate": 1.6466916703209535e-05,
+      "loss": 0.8506,
+      "step": 13596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8659513204098226,
+      "learning_rate": 1.6462996729459594e-05,
+      "loss": 0.8486,
+      "step": 13597
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0405553764691557,
+      "learning_rate": 1.6459076993290967e-05,
+      "loss": 1.0625,
+      "step": 13598
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.123363054477294,
+      "learning_rate": 1.645515749481276e-05,
+      "loss": 0.8184,
+      "step": 13599
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3701181934356126,
+      "learning_rate": 1.6451238234134042e-05,
+      "loss": 0.9004,
+      "step": 13600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4082589618678867,
+      "learning_rate": 1.644731921136388e-05,
+      "loss": 0.9844,
+      "step": 13601
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.2323432355263515,
+      "learning_rate": 1.6443400426611348e-05,
+      "loss": 1.0762,
+      "step": 13602
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.200047200570089,
+      "learning_rate": 1.6439481879985496e-05,
+      "loss": 1.1758,
+      "step": 13603
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.42466013976101,
+      "learning_rate": 1.643556357159538e-05,
+      "loss": 1.1953,
+      "step": 13604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.465697448908812,
+      "learning_rate": 1.643164550155004e-05,
+      "loss": 0.9062,
+      "step": 13605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.036963021029609,
+      "learning_rate": 1.642772766995852e-05,
+      "loss": 0.9258,
+      "step": 13606
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5144058021594984,
+      "learning_rate": 1.6423810076929847e-05,
+      "loss": 0.8584,
+      "step": 13607
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.9450660391043835,
+      "learning_rate": 1.6419892722573043e-05,
+      "loss": 1.9219,
+      "step": 13608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.324612008812678,
+      "learning_rate": 1.6415975606997137e-05,
+      "loss": 0.877,
+      "step": 13609
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.367705033637479,
+      "learning_rate": 1.6412058730311136e-05,
+      "loss": 1.3281,
+      "step": 13610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.277015518430451,
+      "learning_rate": 1.6408142092624043e-05,
+      "loss": 0.8867,
+      "step": 13611
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.114334605454402,
+      "learning_rate": 1.640422569404486e-05,
+      "loss": 0.9497,
+      "step": 13612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.088876463991976,
+      "learning_rate": 1.640030953468258e-05,
+      "loss": 0.9258,
+      "step": 13613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.091577456067603,
+      "learning_rate": 1.639639361464618e-05,
+      "loss": 1.1426,
+      "step": 13614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5452477809295706,
+      "learning_rate": 1.6392477934044644e-05,
+      "loss": 0.877,
+      "step": 13615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.302936220911871,
+      "learning_rate": 1.638856249298694e-05,
+      "loss": 1.3643,
+      "step": 13616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0730164342542565,
+      "learning_rate": 1.6384647291582033e-05,
+      "loss": 0.9971,
+      "step": 13617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5715682374021265,
+      "learning_rate": 1.638073232993889e-05,
+      "loss": 0.9492,
+      "step": 13618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1047117326753164,
+      "learning_rate": 1.6376817608166463e-05,
+      "loss": 0.874,
+      "step": 13619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9616633545815296,
+      "learning_rate": 1.6372903126373686e-05,
+      "loss": 0.5317,
+      "step": 13620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8565572336684362,
+      "learning_rate": 1.6368988884669504e-05,
+      "loss": 0.6206,
+      "step": 13621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.405091694841984,
+      "learning_rate": 1.6365074883162845e-05,
+      "loss": 0.8467,
+      "step": 13622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.0499382997275015,
+      "learning_rate": 1.636116112196264e-05,
+      "loss": 0.8213,
+      "step": 13623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0507156032869847,
+      "learning_rate": 1.63572476011778e-05,
+      "loss": 1.3057,
+      "step": 13624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7576731357536244,
+      "learning_rate": 1.6353334320917247e-05,
+      "loss": 0.7725,
+      "step": 13625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.885586362914289,
+      "learning_rate": 1.634942128128986e-05,
+      "loss": 1.0762,
+      "step": 13626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7849345330379613,
+      "learning_rate": 1.6345508482404582e-05,
+      "loss": 0.6405,
+      "step": 13627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.781854124123497,
+      "learning_rate": 1.6341595924370266e-05,
+      "loss": 1.1699,
+      "step": 13628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1100619530103417,
+      "learning_rate": 1.6337683607295818e-05,
+      "loss": 1.2773,
+      "step": 13629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.830046253373256,
+      "learning_rate": 1.6333771531290103e-05,
+      "loss": 1.0391,
+      "step": 13630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0256811244478574,
+      "learning_rate": 1.6329859696462008e-05,
+      "loss": 1.1123,
+      "step": 13631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.762102410425593,
+      "learning_rate": 1.632594810292038e-05,
+      "loss": 1.3066,
+      "step": 13632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.005477672828417,
+      "learning_rate": 1.632203675077409e-05,
+      "loss": 0.9317,
+      "step": 13633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.131467506547416,
+      "learning_rate": 1.6318125640131977e-05,
+      "loss": 0.9336,
+      "step": 13634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9314217049318527,
+      "learning_rate": 1.631421477110289e-05,
+      "loss": 1.207,
+      "step": 13635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8754018352978385,
+      "learning_rate": 1.631030414379568e-05,
+      "loss": 1.0498,
+      "step": 13636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.444679946095177,
+      "learning_rate": 1.6306393758319163e-05,
+      "loss": 0.957,
+      "step": 13637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.7936656171090615,
+      "learning_rate": 1.6302483614782173e-05,
+      "loss": 0.9736,
+      "step": 13638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.943796320142417,
+      "learning_rate": 1.629857371329352e-05,
+      "loss": 1.0723,
+      "step": 13639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5678974156790555,
+      "learning_rate": 1.6294664053962023e-05,
+      "loss": 1.2734,
+      "step": 13640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3227141310546227,
+      "learning_rate": 1.6290754636896478e-05,
+      "loss": 1.2007,
+      "step": 13641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.064306797121693,
+      "learning_rate": 1.6286845462205684e-05,
+      "loss": 0.9746,
+      "step": 13642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.215580462827284,
+      "learning_rate": 1.6282936529998434e-05,
+      "loss": 1.4512,
+      "step": 13643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1742479309000355,
+      "learning_rate": 1.6279027840383517e-05,
+      "loss": 1.4453,
+      "step": 13644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.856137488202047,
+      "learning_rate": 1.6275119393469692e-05,
+      "loss": 1.0234,
+      "step": 13645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2872201842218995,
+      "learning_rate": 1.627121118936575e-05,
+      "loss": 1.0996,
+      "step": 13646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9055843128979157,
+      "learning_rate": 1.6267303228180448e-05,
+      "loss": 0.6106,
+      "step": 13647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.43586442920125,
+      "learning_rate": 1.6263395510022543e-05,
+      "loss": 1.5273,
+      "step": 13648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.373355880166229,
+      "learning_rate": 1.625948803500078e-05,
+      "loss": 0.3591,
+      "step": 13649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.823588489076767,
+      "learning_rate": 1.6255580803223915e-05,
+      "loss": 1.1992,
+      "step": 13650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7246986062413727,
+      "learning_rate": 1.6251673814800672e-05,
+      "loss": 1.2461,
+      "step": 13651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5153129810622366,
+      "learning_rate": 1.6247767069839788e-05,
+      "loss": 0.6755,
+      "step": 13652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.539168729422556,
+      "learning_rate": 1.6243860568449977e-05,
+      "loss": 0.8408,
+      "step": 13653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.648905664886174,
+      "learning_rate": 1.6239954310739957e-05,
+      "loss": 1.1445,
+      "step": 13654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.195838626875558,
+      "learning_rate": 1.6236048296818454e-05,
+      "loss": 1.2617,
+      "step": 13655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8435967372633564,
+      "learning_rate": 1.6232142526794152e-05,
+      "loss": 0.7686,
+      "step": 13656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9360420992065133,
+      "learning_rate": 1.6228237000775763e-05,
+      "loss": 1.4062,
+      "step": 13657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8899188854621474,
+      "learning_rate": 1.622433171887196e-05,
+      "loss": 1.1309,
+      "step": 13658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.318974581225055,
+      "learning_rate": 1.6220426681191437e-05,
+      "loss": 1.1826,
+      "step": 13659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1902847003955035,
+      "learning_rate": 1.6216521887842863e-05,
+      "loss": 0.9062,
+      "step": 13660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2859735576004776,
+      "learning_rate": 1.621261733893491e-05,
+      "loss": 1.1172,
+      "step": 13661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.957274615098206,
+      "learning_rate": 1.6208713034576232e-05,
+      "loss": 1.4707,
+      "step": 13662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7999470199629382,
+      "learning_rate": 1.6204808974875494e-05,
+      "loss": 0.6006,
+      "step": 13663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.743822537213964,
+      "learning_rate": 1.620090515994135e-05,
+      "loss": 1.626,
+      "step": 13664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5831339728913276,
+      "learning_rate": 1.6197001589882425e-05,
+      "loss": 0.7588,
+      "step": 13665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.025738442553821,
+      "learning_rate": 1.6193098264807362e-05,
+      "loss": 0.7495,
+      "step": 13666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.226454133190744,
+      "learning_rate": 1.6189195184824795e-05,
+      "loss": 0.8818,
+      "step": 13667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1230355764140523,
+      "learning_rate": 1.6185292350043335e-05,
+      "loss": 1.0361,
+      "step": 13668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3182201263336024,
+      "learning_rate": 1.6181389760571606e-05,
+      "loss": 0.9609,
+      "step": 13669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5072131080239783,
+      "learning_rate": 1.6177487416518206e-05,
+      "loss": 1.167,
+      "step": 13670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1283716492974056,
+      "learning_rate": 1.617358531799174e-05,
+      "loss": 1.1094,
+      "step": 13671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.486961974297204,
+      "learning_rate": 1.61696834651008e-05,
+      "loss": 1.1387,
+      "step": 13672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2283304256210053,
+      "learning_rate": 1.6165781857953965e-05,
+      "loss": 0.6037,
+      "step": 13673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5113579519101825,
+      "learning_rate": 1.616188049665984e-05,
+      "loss": 1.2266,
+      "step": 13674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.787253731746625,
+      "learning_rate": 1.6157979381326975e-05,
+      "loss": 1.9121,
+      "step": 13675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.615325629855105,
+      "learning_rate": 1.615407851206395e-05,
+      "loss": 0.9814,
+      "step": 13676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.586550574315631,
+      "learning_rate": 1.6150177888979313e-05,
+      "loss": 1.8613,
+      "step": 13677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6844791911713584,
+      "learning_rate": 1.6146277512181628e-05,
+      "loss": 0.6611,
+      "step": 13678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.599587252585642,
+      "learning_rate": 1.614237738177943e-05,
+      "loss": 0.6934,
+      "step": 13679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.455113613997969,
+      "learning_rate": 1.6138477497881272e-05,
+      "loss": 1.2324,
+      "step": 13680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9728452043469717,
+      "learning_rate": 1.6134577860595666e-05,
+      "loss": 0.8037,
+      "step": 13681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.909061835425487,
+      "learning_rate": 1.613067847003115e-05,
+      "loss": 1.3398,
+      "step": 13682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3844162308848706,
+      "learning_rate": 1.612677932629625e-05,
+      "loss": 0.8887,
+      "step": 13683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9104118490941544,
+      "learning_rate": 1.6122880429499463e-05,
+      "loss": 1.0107,
+      "step": 13684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1889113096948916,
+      "learning_rate": 1.6118981779749303e-05,
+      "loss": 1.0352,
+      "step": 13685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.656500073572496,
+      "learning_rate": 1.6115083377154265e-05,
+      "loss": 0.999,
+      "step": 13686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3870062272526376,
+      "learning_rate": 1.6111185221822833e-05,
+      "loss": 0.8613,
+      "step": 13687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.913188404374548,
+      "learning_rate": 1.6107287313863507e-05,
+      "loss": 1.0527,
+      "step": 13688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.162337183473414,
+      "learning_rate": 1.6103389653384748e-05,
+      "loss": 1.2939,
+      "step": 13689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.105695887258079,
+      "learning_rate": 1.609949224049504e-05,
+      "loss": 0.9375,
+      "step": 13690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7695624409804493,
+      "learning_rate": 1.6095595075302827e-05,
+      "loss": 0.6133,
+      "step": 13691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.956110009697389,
+      "learning_rate": 1.6091698157916586e-05,
+      "loss": 1.1685,
+      "step": 13692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.181328159451275,
+      "learning_rate": 1.6087801488444762e-05,
+      "loss": 0.7119,
+      "step": 13693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2826483634203036,
+      "learning_rate": 1.608390506699579e-05,
+      "loss": 1.0303,
+      "step": 13694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4182368362527513,
+      "learning_rate": 1.6080008893678117e-05,
+      "loss": 1.2217,
+      "step": 13695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.819805418449139,
+      "learning_rate": 1.6076112968600156e-05,
+      "loss": 0.7012,
+      "step": 13696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.797092450602156,
+      "learning_rate": 1.6072217291870343e-05,
+      "loss": 1.1777,
+      "step": 13697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0451288127668943,
+      "learning_rate": 1.6068321863597087e-05,
+      "loss": 1.207,
+      "step": 13698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.806114022225119,
+      "learning_rate": 1.60644266838888e-05,
+      "loss": 1.8516,
+      "step": 13699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2118557176309936,
+      "learning_rate": 1.6060531752853876e-05,
+      "loss": 1.375,
+      "step": 13700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4768633253624706,
+      "learning_rate": 1.605663707060071e-05,
+      "loss": 1.0791,
+      "step": 13701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4076586968906715,
+      "learning_rate": 1.6052742637237707e-05,
+      "loss": 0.4668,
+      "step": 13702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.574250706047783,
+      "learning_rate": 1.6048848452873222e-05,
+      "loss": 0.6169,
+      "step": 13703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.209008659411442,
+      "learning_rate": 1.604495451761565e-05,
+      "loss": 0.6763,
+      "step": 13704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8020725651833107,
+      "learning_rate": 1.604106083157334e-05,
+      "loss": 1.2227,
+      "step": 13705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.801077444016709,
+      "learning_rate": 1.6037167394854663e-05,
+      "loss": 0.6841,
+      "step": 13706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7594565366050663,
+      "learning_rate": 1.603327420756797e-05,
+      "loss": 0.8301,
+      "step": 13707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.15383476476349,
+      "learning_rate": 1.6029381269821604e-05,
+      "loss": 1.1211,
+      "step": 13708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.362768827561905,
+      "learning_rate": 1.602548858172391e-05,
+      "loss": 0.8906,
+      "step": 13709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.229648016787444,
+      "learning_rate": 1.6021596143383205e-05,
+      "loss": 1.5625,
+      "step": 13710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.789606736207408,
+      "learning_rate": 1.6017703954907827e-05,
+      "loss": 1.1914,
+      "step": 13711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.475003302215531,
+      "learning_rate": 1.60138120164061e-05,
+      "loss": 1.0674,
+      "step": 13712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3828694321131545,
+      "learning_rate": 1.6009920327986318e-05,
+      "loss": 1.4453,
+      "step": 13713
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.473574721469869,
+      "learning_rate": 1.6006028889756803e-05,
+      "loss": 1.0127,
+      "step": 13714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6993060173799983,
+      "learning_rate": 1.600213770182583e-05,
+      "loss": 0.4934,
+      "step": 13715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6259402453094842,
+      "learning_rate": 1.5998246764301713e-05,
+      "loss": 1.0557,
+      "step": 13716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.189102022194082,
+      "learning_rate": 1.5994356077292715e-05,
+      "loss": 0.8916,
+      "step": 13717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.101129535271181,
+      "learning_rate": 1.599046564090713e-05,
+      "loss": 1.5801,
+      "step": 13718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.008544650218964,
+      "learning_rate": 1.5986575455253204e-05,
+      "loss": 1.4307,
+      "step": 13719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1519082627508,
+      "learning_rate": 1.5982685520439222e-05,
+      "loss": 0.6648,
+      "step": 13720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4189445239956324,
+      "learning_rate": 1.5978795836573435e-05,
+      "loss": 0.981,
+      "step": 13721
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.857921126497931,
+      "learning_rate": 1.5974906403764084e-05,
+      "loss": 0.4956,
+      "step": 13722
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.565754010654271,
+      "learning_rate": 1.5971017222119418e-05,
+      "loss": 1.002,
+      "step": 13723
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.980689333350512,
+      "learning_rate": 1.5967128291747656e-05,
+      "loss": 1.1035,
+      "step": 13724
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.22292507696712,
+      "learning_rate": 1.596323961275705e-05,
+      "loss": 1.4199,
+      "step": 13725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5912606116453434,
+      "learning_rate": 1.59593511852558e-05,
+      "loss": 0.9937,
+      "step": 13726
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7559999151016568,
+      "learning_rate": 1.5955463009352123e-05,
+      "loss": 0.7783,
+      "step": 13727
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.192411360875603,
+      "learning_rate": 1.5951575085154224e-05,
+      "loss": 1.2109,
+      "step": 13728
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.735249204520299,
+      "learning_rate": 1.5947687412770317e-05,
+      "loss": 1.1846,
+      "step": 13729
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5909225495458874,
+      "learning_rate": 1.5943799992308576e-05,
+      "loss": 1.3867,
+      "step": 13730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.734972015818107,
+      "learning_rate": 1.5939912823877207e-05,
+      "loss": 1.2031,
+      "step": 13731
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4915124979026286,
+      "learning_rate": 1.5936025907584368e-05,
+      "loss": 1.209,
+      "step": 13732
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8317577992982943,
+      "learning_rate": 1.593213924353824e-05,
+      "loss": 0.8721,
+      "step": 13733
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.452600085377445,
+      "learning_rate": 1.592825283184698e-05,
+      "loss": 1.4141,
+      "step": 13734
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5782036046847567,
+      "learning_rate": 1.592436667261876e-05,
+      "loss": 0.8916,
+      "step": 13735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.982050938972714,
+      "learning_rate": 1.5920480765961716e-05,
+      "loss": 0.7473,
+      "step": 13736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0645808721481043,
+      "learning_rate": 1.5916595111983985e-05,
+      "loss": 1.0039,
+      "step": 13737
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.703322386839028,
+      "learning_rate": 1.5912709710793727e-05,
+      "loss": 0.9573,
+      "step": 13738
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.662031509588862,
+      "learning_rate": 1.5908824562499054e-05,
+      "loss": 0.6572,
+      "step": 13739
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.753190023465286,
+      "learning_rate": 1.59049396672081e-05,
+      "loss": 1.498,
+      "step": 13740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.275964964062312,
+      "learning_rate": 1.5901055025028963e-05,
+      "loss": 0.6416,
+      "step": 13741
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.666253226734754,
+      "learning_rate": 1.5897170636069768e-05,
+      "loss": 1.001,
+      "step": 13742
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0879398851126982,
+      "learning_rate": 1.5893286500438603e-05,
+      "loss": 1.1133,
+      "step": 13743
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1479667009013785,
+      "learning_rate": 1.588940261824357e-05,
+      "loss": 1.2871,
+      "step": 13744
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.362355297060634,
+      "learning_rate": 1.588551898959275e-05,
+      "loss": 0.8379,
+      "step": 13745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0835654955887417,
+      "learning_rate": 1.5881635614594228e-05,
+      "loss": 0.635,
+      "step": 13746
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1506548172674966,
+      "learning_rate": 1.587775249335607e-05,
+      "loss": 0.8281,
+      "step": 13747
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2058879566092187,
+      "learning_rate": 1.5873869625986353e-05,
+      "loss": 1.418,
+      "step": 13748
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6645318565933915,
+      "learning_rate": 1.5869987012593123e-05,
+      "loss": 1.0078,
+      "step": 13749
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4462713520549464,
+      "learning_rate": 1.586610465328445e-05,
+      "loss": 1.1602,
+      "step": 13750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7763934267191739,
+      "learning_rate": 1.5862222548168355e-05,
+      "loss": 1.1445,
+      "step": 13751
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4670736681604906,
+      "learning_rate": 1.585834069735289e-05,
+      "loss": 1.084,
+      "step": 13752
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.152380952268474,
+      "learning_rate": 1.585445910094608e-05,
+      "loss": 1.4287,
+      "step": 13753
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8289418148215004,
+      "learning_rate": 1.5850577759055956e-05,
+      "loss": 1.2441,
+      "step": 13754
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1799358299066696,
+      "learning_rate": 1.5846696671790522e-05,
+      "loss": 0.5052,
+      "step": 13755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.347429403059299,
+      "learning_rate": 1.584281583925779e-05,
+      "loss": 1.3281,
+      "step": 13756
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.922885295271829,
+      "learning_rate": 1.583893526156577e-05,
+      "loss": 1.5684,
+      "step": 13757
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.517121399199428,
+      "learning_rate": 1.5835054938822453e-05,
+      "loss": 0.8801,
+      "step": 13758
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5334097984552715,
+      "learning_rate": 1.583117487113583e-05,
+      "loss": 0.8403,
+      "step": 13759
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8044146492817306,
+      "learning_rate": 1.582729505861387e-05,
+      "loss": 1.6074,
+      "step": 13760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1715683060624014,
+      "learning_rate": 1.5823415501364557e-05,
+      "loss": 0.998,
+      "step": 13761
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9501030872268688,
+      "learning_rate": 1.5819536199495854e-05,
+      "loss": 0.6143,
+      "step": 13762
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.505079259475125,
+      "learning_rate": 1.5815657153115728e-05,
+      "loss": 1.4707,
+      "step": 13763
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.741153964547596,
+      "learning_rate": 1.581177836233212e-05,
+      "loss": 0.708,
+      "step": 13764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.875821701051335,
+      "learning_rate": 1.5807899827252968e-05,
+      "loss": 1.3672,
+      "step": 13765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.654025576965056,
+      "learning_rate": 1.580402154798623e-05,
+      "loss": 1.0156,
+      "step": 13766
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.663667116610488,
+      "learning_rate": 1.5800143524639836e-05,
+      "loss": 0.8662,
+      "step": 13767
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6660470143241417,
+      "learning_rate": 1.5796265757321693e-05,
+      "loss": 0.5801,
+      "step": 13768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.920263073131832,
+      "learning_rate": 1.579238824613974e-05,
+      "loss": 0.8955,
+      "step": 13769
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7448653923619424,
+      "learning_rate": 1.5788510991201862e-05,
+      "loss": 1.3203,
+      "step": 13770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4113552087625765,
+      "learning_rate": 1.5784633992615978e-05,
+      "loss": 0.9502,
+      "step": 13771
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.799999289677017,
+      "learning_rate": 1.5780757250489976e-05,
+      "loss": 0.6953,
+      "step": 13772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.6182980016557655,
+      "learning_rate": 1.577688076493175e-05,
+      "loss": 1.3242,
+      "step": 13773
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1369084303662866,
+      "learning_rate": 1.577300453604917e-05,
+      "loss": 0.959,
+      "step": 13774
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4923435662746445,
+      "learning_rate": 1.576912856395012e-05,
+      "loss": 0.874,
+      "step": 13775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0126932987238226,
+      "learning_rate": 1.5765252848742468e-05,
+      "loss": 1.0039,
+      "step": 13776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3959419115313745,
+      "learning_rate": 1.5761377390534066e-05,
+      "loss": 0.3477,
+      "step": 13777
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6459842448612494,
+      "learning_rate": 1.5757502189432778e-05,
+      "loss": 1.0762,
+      "step": 13778
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.234907291348265,
+      "learning_rate": 1.575362724554643e-05,
+      "loss": 1.7793,
+      "step": 13779
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5673752904689637,
+      "learning_rate": 1.5749752558982885e-05,
+      "loss": 0.9585,
+      "step": 13780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.493336953587207,
+      "learning_rate": 1.5745878129849952e-05,
+      "loss": 1.2656,
+      "step": 13781
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.076120772499908,
+      "learning_rate": 1.5742003958255462e-05,
+      "loss": 0.6587,
+      "step": 13782
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.947254299700739,
+      "learning_rate": 1.5738130044307236e-05,
+      "loss": 1.1445,
+      "step": 13783
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.690358879155513,
+      "learning_rate": 1.5734256388113074e-05,
+      "loss": 1.6221,
+      "step": 13784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6438915205789195,
+      "learning_rate": 1.5730382989780786e-05,
+      "loss": 1.0596,
+      "step": 13785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.911293674067296,
+      "learning_rate": 1.5726509849418176e-05,
+      "loss": 0.7144,
+      "step": 13786
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0259890176960798,
+      "learning_rate": 1.5722636967133013e-05,
+      "loss": 1.3379,
+      "step": 13787
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.74290991502472,
+      "learning_rate": 1.5718764343033095e-05,
+      "loss": 1.1768,
+      "step": 13788
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1707048935677893,
+      "learning_rate": 1.5714891977226177e-05,
+      "loss": 1.3203,
+      "step": 13789
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4060628248464657,
+      "learning_rate": 1.5711019869820042e-05,
+      "loss": 1.0713,
+      "step": 13790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8661643549238476,
+      "learning_rate": 1.5707148020922434e-05,
+      "loss": 1.5391,
+      "step": 13791
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5439190707619663,
+      "learning_rate": 1.5703276430641125e-05,
+      "loss": 0.9424,
+      "step": 13792
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.890250186848558,
+      "learning_rate": 1.5699405099083835e-05,
+      "loss": 1.1182,
+      "step": 13793
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.091679173053551,
+      "learning_rate": 1.5695534026358322e-05,
+      "loss": 1.1494,
+      "step": 13794
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7376803956276403,
+      "learning_rate": 1.569166321257231e-05,
+      "loss": 1.6328,
+      "step": 13795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7766041045756134,
+      "learning_rate": 1.5687792657833523e-05,
+      "loss": 1.1641,
+      "step": 13796
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.472330455093312,
+      "learning_rate": 1.5683922362249677e-05,
+      "loss": 1.4609,
+      "step": 13797
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.890386128221426,
+      "learning_rate": 1.568005232592847e-05,
+      "loss": 1.5645,
+      "step": 13798
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8436715251449,
+      "learning_rate": 1.5676182548977623e-05,
+      "loss": 0.7881,
+      "step": 13799
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.703190155099938,
+      "learning_rate": 1.567231303150481e-05,
+      "loss": 0.8252,
+      "step": 13800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9428176508223627,
+      "learning_rate": 1.5668443773617737e-05,
+      "loss": 1.0991,
+      "step": 13801
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0456687691445983,
+      "learning_rate": 1.5664574775424058e-05,
+      "loss": 0.5386,
+      "step": 13802
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.018990361221379,
+      "learning_rate": 1.566070603703148e-05,
+      "loss": 1.2061,
+      "step": 13803
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.601871063424818,
+      "learning_rate": 1.5656837558547644e-05,
+      "loss": 1.1221,
+      "step": 13804
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9677163695586446,
+      "learning_rate": 1.565296934008022e-05,
+      "loss": 1.1562,
+      "step": 13805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6764719718949257,
+      "learning_rate": 1.564910138173685e-05,
+      "loss": 0.9209,
+      "step": 13806
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8491212192721695,
+      "learning_rate": 1.5645233683625188e-05,
+      "loss": 1.1211,
+      "step": 13807
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.032138696934107,
+      "learning_rate": 1.5641366245852857e-05,
+      "loss": 1.4297,
+      "step": 13808
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.361240429326038,
+      "learning_rate": 1.5637499068527498e-05,
+      "loss": 1.4316,
+      "step": 13809
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.503941099784291,
+      "learning_rate": 1.5633632151756723e-05,
+      "loss": 1.4863,
+      "step": 13810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0009957886119403,
+      "learning_rate": 1.562976549564816e-05,
+      "loss": 0.9141,
+      "step": 13811
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.113719958109355,
+      "learning_rate": 1.56258991003094e-05,
+      "loss": 0.7292,
+      "step": 13812
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9528286901299228,
+      "learning_rate": 1.5622032965848052e-05,
+      "loss": 1.1504,
+      "step": 13813
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.44609364473336,
+      "learning_rate": 1.5618167092371716e-05,
+      "loss": 1.4941,
+      "step": 13814
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.648604292241353,
+      "learning_rate": 1.5614301479987964e-05,
+      "loss": 1.0781,
+      "step": 13815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.620338114986886,
+      "learning_rate": 1.561043612880439e-05,
+      "loss": 1.2109,
+      "step": 13816
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.615875109260648,
+      "learning_rate": 1.5606571038928548e-05,
+      "loss": 1.2793,
+      "step": 13817
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6842912840776005,
+      "learning_rate": 1.5602706210468015e-05,
+      "loss": 0.8896,
+      "step": 13818
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.404543382912868,
+      "learning_rate": 1.5598841643530337e-05,
+      "loss": 0.7197,
+      "step": 13819
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9108295908910757,
+      "learning_rate": 1.5594977338223077e-05,
+      "loss": 0.9062,
+      "step": 13820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.829838026027996,
+      "learning_rate": 1.5591113294653753e-05,
+      "loss": 1.3735,
+      "step": 13821
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.478193353814258,
+      "learning_rate": 1.5587249512929925e-05,
+      "loss": 0.998,
+      "step": 13822
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.259869189300033,
+      "learning_rate": 1.558338599315911e-05,
+      "loss": 0.8105,
+      "step": 13823
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8745424984041745,
+      "learning_rate": 1.5579522735448837e-05,
+      "loss": 1.1582,
+      "step": 13824
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.137826133966709,
+      "learning_rate": 1.5575659739906602e-05,
+      "loss": 0.9648,
+      "step": 13825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9908841235111394,
+      "learning_rate": 1.5571797006639926e-05,
+      "loss": 0.8535,
+      "step": 13826
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.965961282485168,
+      "learning_rate": 1.5567934535756298e-05,
+      "loss": 1.8945,
+      "step": 13827
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.375815844222868,
+      "learning_rate": 1.5564072327363217e-05,
+      "loss": 1.2217,
+      "step": 13828
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 22.061638229008658,
+      "learning_rate": 1.5560210381568154e-05,
+      "loss": 1.541,
+      "step": 13829
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8328523057627106,
+      "learning_rate": 1.555634869847859e-05,
+      "loss": 0.6279,
+      "step": 13830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3845147129572535,
+      "learning_rate": 1.5552487278202007e-05,
+      "loss": 1.3516,
+      "step": 13831
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.767601032889468,
+      "learning_rate": 1.5548626120845852e-05,
+      "loss": 1.2285,
+      "step": 13832
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.398203192252921,
+      "learning_rate": 1.5544765226517588e-05,
+      "loss": 0.5867,
+      "step": 13833
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.219768559016215,
+      "learning_rate": 1.5540904595324655e-05,
+      "loss": 0.814,
+      "step": 13834
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6448901340043163,
+      "learning_rate": 1.5537044227374503e-05,
+      "loss": 1.3242,
+      "step": 13835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7141423259233575,
+      "learning_rate": 1.553318412277455e-05,
+      "loss": 1.0898,
+      "step": 13836
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.480660823912345,
+      "learning_rate": 1.5529324281632234e-05,
+      "loss": 0.9775,
+      "step": 13837
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.480439163549328,
+      "learning_rate": 1.5525464704054965e-05,
+      "loss": 1.1387,
+      "step": 13838
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.189318309924441,
+      "learning_rate": 1.552160539015015e-05,
+      "loss": 0.8884,
+      "step": 13839
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1206385885745385,
+      "learning_rate": 1.5517746340025206e-05,
+      "loss": 0.9189,
+      "step": 13840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6375175113910214,
+      "learning_rate": 1.551388755378752e-05,
+      "loss": 0.8535,
+      "step": 13841
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.651578407936729,
+      "learning_rate": 1.5510029031544478e-05,
+      "loss": 0.8994,
+      "step": 13842
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8666435311223415,
+      "learning_rate": 1.5506170773403472e-05,
+      "loss": 1.2109,
+      "step": 13843
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7832876972913483,
+      "learning_rate": 1.5502312779471863e-05,
+      "loss": 0.6748,
+      "step": 13844
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.286974582886162,
+      "learning_rate": 1.549845504985703e-05,
+      "loss": 1.0977,
+      "step": 13845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.3376661549165805,
+      "learning_rate": 1.5494597584666315e-05,
+      "loss": 1.5625,
+      "step": 13846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4651667327178743,
+      "learning_rate": 1.549074038400709e-05,
+      "loss": 1.0586,
+      "step": 13847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3780189010515937,
+      "learning_rate": 1.548688344798668e-05,
+      "loss": 0.811,
+      "step": 13848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.833729166589927,
+      "learning_rate": 1.548302677671243e-05,
+      "loss": 1.2891,
+      "step": 13849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1451506389693136,
+      "learning_rate": 1.5479170370291675e-05,
+      "loss": 1.375,
+      "step": 13850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7540151082112367,
+      "learning_rate": 1.5475314228831726e-05,
+      "loss": 1.5703,
+      "step": 13851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9685537452366222,
+      "learning_rate": 1.5471458352439917e-05,
+      "loss": 0.4536,
+      "step": 13852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.932351066015788,
+      "learning_rate": 1.5467602741223536e-05,
+      "loss": 1.2676,
+      "step": 13853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7729155995074002,
+      "learning_rate": 1.546374739528989e-05,
+      "loss": 1.2129,
+      "step": 13854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6477503602520467,
+      "learning_rate": 1.545989231474627e-05,
+      "loss": 1.5586,
+      "step": 13855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.763521724647785,
+      "learning_rate": 1.5456037499699972e-05,
+      "loss": 0.9424,
+      "step": 13856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.359402536678566,
+      "learning_rate": 1.5452182950258252e-05,
+      "loss": 1.2188,
+      "step": 13857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7166564597969765,
+      "learning_rate": 1.5448328666528396e-05,
+      "loss": 1.0703,
+      "step": 13858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.096181277573591,
+      "learning_rate": 1.5444474648617675e-05,
+      "loss": 0.9111,
+      "step": 13859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.0567814412698455,
+      "learning_rate": 1.5440620896633324e-05,
+      "loss": 0.9736,
+      "step": 13860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.03758753489054,
+      "learning_rate": 1.5436767410682607e-05,
+      "loss": 1.2554,
+      "step": 13861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9209804046538155,
+      "learning_rate": 1.5432914190872757e-05,
+      "loss": 0.8115,
+      "step": 13862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0645690468135953,
+      "learning_rate": 1.542906123731101e-05,
+      "loss": 0.9092,
+      "step": 13863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9011375826874604,
+      "learning_rate": 1.5425208550104597e-05,
+      "loss": 1.0488,
+      "step": 13864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.325777957338585,
+      "learning_rate": 1.5421356129360733e-05,
+      "loss": 1.0371,
+      "step": 13865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7382292664716594,
+      "learning_rate": 1.5417503975186624e-05,
+      "loss": 1.0283,
+      "step": 13866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5294547943388277,
+      "learning_rate": 1.5413652087689472e-05,
+      "loss": 0.6698,
+      "step": 13867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0561012840363415,
+      "learning_rate": 1.5409800466976487e-05,
+      "loss": 1.1807,
+      "step": 13868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.677697958444704,
+      "learning_rate": 1.5405949113154855e-05,
+      "loss": 1.0273,
+      "step": 13869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4885949336839692,
+      "learning_rate": 1.5402098026331744e-05,
+      "loss": 1.146,
+      "step": 13870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.081827826683878,
+      "learning_rate": 1.539824720661435e-05,
+      "loss": 1.2754,
+      "step": 13871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4596083185020206,
+      "learning_rate": 1.5394396654109815e-05,
+      "loss": 0.6797,
+      "step": 13872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8882654840856024,
+      "learning_rate": 1.5390546368925322e-05,
+      "loss": 1.1196,
+      "step": 13873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.353485132550134,
+      "learning_rate": 1.5386696351168003e-05,
+      "loss": 1.2227,
+      "step": 13874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.407149704021501,
+      "learning_rate": 1.5382846600945018e-05,
+      "loss": 1.4727,
+      "step": 13875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.801994624854956,
+      "learning_rate": 1.5378997118363488e-05,
+      "loss": 1.0972,
+      "step": 13876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.738102056759386,
+      "learning_rate": 1.5375147903530553e-05,
+      "loss": 1.0645,
+      "step": 13877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8690374708217234,
+      "learning_rate": 1.5371298956553336e-05,
+      "loss": 0.7993,
+      "step": 13878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1195050044953447,
+      "learning_rate": 1.536745027753895e-05,
+      "loss": 1.4043,
+      "step": 13879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.352075670353078,
+      "learning_rate": 1.5363601866594508e-05,
+      "loss": 1.3008,
+      "step": 13880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.343458644259202,
+      "learning_rate": 1.5359753723827092e-05,
+      "loss": 0.9053,
+      "step": 13881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.369326874380962,
+      "learning_rate": 1.535590584934381e-05,
+      "loss": 0.7952,
+      "step": 13882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2642022236528976,
+      "learning_rate": 1.5352058243251745e-05,
+      "loss": 0.6799,
+      "step": 13883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.661959353155474,
+      "learning_rate": 1.534821090565796e-05,
+      "loss": 1.4473,
+      "step": 13884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0736554997931522,
+      "learning_rate": 1.534436383666955e-05,
+      "loss": 0.917,
+      "step": 13885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.01652712652125,
+      "learning_rate": 1.5340517036393553e-05,
+      "loss": 0.9902,
+      "step": 13886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0289568534994564,
+      "learning_rate": 1.5336670504937038e-05,
+      "loss": 0.7803,
+      "step": 13887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2047942487401295,
+      "learning_rate": 1.5332824242407052e-05,
+      "loss": 1.1914,
+      "step": 13888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0545507532131593,
+      "learning_rate": 1.532897824891063e-05,
+      "loss": 1.1641,
+      "step": 13889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8493405755819126,
+      "learning_rate": 1.532513252455481e-05,
+      "loss": 1.3125,
+      "step": 13890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6486301412103463,
+      "learning_rate": 1.5321287069446605e-05,
+      "loss": 0.7925,
+      "step": 13891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.112546105417789,
+      "learning_rate": 1.531744188369305e-05,
+      "loss": 0.858,
+      "step": 13892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.436036396706548,
+      "learning_rate": 1.5313596967401138e-05,
+      "loss": 1.6992,
+      "step": 13893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.182832387449596,
+      "learning_rate": 1.5309752320677884e-05,
+      "loss": 1.3008,
+      "step": 13894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.8147898589136835,
+      "learning_rate": 1.530590794363027e-05,
+      "loss": 1.3359,
+      "step": 13895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.9048168251606405,
+      "learning_rate": 1.5302063836365295e-05,
+      "loss": 0.7046,
+      "step": 13896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4045855227257555,
+      "learning_rate": 1.5298219998989944e-05,
+      "loss": 0.6597,
+      "step": 13897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.702119587104828,
+      "learning_rate": 1.5294376431611173e-05,
+      "loss": 1.2832,
+      "step": 13898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.373560820760368,
+      "learning_rate": 1.529053313433596e-05,
+      "loss": 0.6543,
+      "step": 13899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.148180937511288,
+      "learning_rate": 1.528669010727125e-05,
+      "loss": 0.4458,
+      "step": 13900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5233155629104767,
+      "learning_rate": 1.5282847350524006e-05,
+      "loss": 1.3613,
+      "step": 13901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9555191015267934,
+      "learning_rate": 1.527900486420116e-05,
+      "loss": 1.5176,
+      "step": 13902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.229757062202538,
+      "learning_rate": 1.5275162648409652e-05,
+      "loss": 1.2461,
+      "step": 13903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.842535241545172,
+      "learning_rate": 1.5271320703256405e-05,
+      "loss": 1.1084,
+      "step": 13904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.209324043857436,
+      "learning_rate": 1.5267479028848348e-05,
+      "loss": 1.2207,
+      "step": 13905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.629148118021715,
+      "learning_rate": 1.5263637625292385e-05,
+      "loss": 0.9873,
+      "step": 13906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9046916532901486,
+      "learning_rate": 1.5259796492695428e-05,
+      "loss": 1.4121,
+      "step": 13907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0858958950965563,
+      "learning_rate": 1.5255955631164364e-05,
+      "loss": 1.3574,
+      "step": 13908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.303101067834644,
+      "learning_rate": 1.5252115040806092e-05,
+      "loss": 0.8394,
+      "step": 13909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9545045889849297,
+      "learning_rate": 1.5248274721727487e-05,
+      "loss": 1.0176,
+      "step": 13910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8994129619410542,
+      "learning_rate": 1.524443467403543e-05,
+      "loss": 1.6133,
+      "step": 13911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4993387301408703,
+      "learning_rate": 1.524059489783678e-05,
+      "loss": 1.1582,
+      "step": 13912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.780657158577991,
+      "learning_rate": 1.5236755393238396e-05,
+      "loss": 1.3145,
+      "step": 13913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.322422927639921,
+      "learning_rate": 1.5232916160347147e-05,
+      "loss": 1.2305,
+      "step": 13914
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.881263621527389,
+      "learning_rate": 1.522907719926986e-05,
+      "loss": 0.9854,
+      "step": 13915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9606872642098403,
+      "learning_rate": 1.5225238510113377e-05,
+      "loss": 0.9016,
+      "step": 13916
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.475146231592744,
+      "learning_rate": 1.5221400092984527e-05,
+      "loss": 1.0918,
+      "step": 13917
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7223517577390677,
+      "learning_rate": 1.5217561947990133e-05,
+      "loss": 0.7417,
+      "step": 13918
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.853114931413624,
+      "learning_rate": 1.5213724075237004e-05,
+      "loss": 0.8147,
+      "step": 13919
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.25231499018932,
+      "learning_rate": 1.5209886474831955e-05,
+      "loss": 0.7891,
+      "step": 13920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.275785616534281,
+      "learning_rate": 1.5206049146881776e-05,
+      "loss": 0.5957,
+      "step": 13921
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.407067405688973,
+      "learning_rate": 1.5202212091493268e-05,
+      "loss": 1.2578,
+      "step": 13922
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2898269782856127,
+      "learning_rate": 1.5198375308773188e-05,
+      "loss": 1.1465,
+      "step": 13923
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.236524601596992,
+      "learning_rate": 1.5194538798828354e-05,
+      "loss": 0.8384,
+      "step": 13924
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.058927806296153,
+      "learning_rate": 1.5190702561765504e-05,
+      "loss": 1.2891,
+      "step": 13925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.147643852107577,
+      "learning_rate": 1.5186866597691419e-05,
+      "loss": 1.4199,
+      "step": 13926
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.688818784212771,
+      "learning_rate": 1.5183030906712831e-05,
+      "loss": 1.6074,
+      "step": 13927
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.206547128699193,
+      "learning_rate": 1.5179195488936504e-05,
+      "loss": 0.8008,
+      "step": 13928
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.43597392105846,
+      "learning_rate": 1.5175360344469164e-05,
+      "loss": 1.0488,
+      "step": 13929
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6368296960165396,
+      "learning_rate": 1.5171525473417546e-05,
+      "loss": 0.9297,
+      "step": 13930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.392822367389835,
+      "learning_rate": 1.5167690875888375e-05,
+      "loss": 1.4082,
+      "step": 13931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.195146821772662,
+      "learning_rate": 1.5163856551988353e-05,
+      "loss": 1.0684,
+      "step": 13932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3190479878818544,
+      "learning_rate": 1.5160022501824215e-05,
+      "loss": 0.718,
+      "step": 13933
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.437701310711093,
+      "learning_rate": 1.5156188725502635e-05,
+      "loss": 1.3027,
+      "step": 13934
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3141746643206607,
+      "learning_rate": 1.5152355223130324e-05,
+      "loss": 0.9941,
+      "step": 13935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0600329991973547,
+      "learning_rate": 1.5148521994813952e-05,
+      "loss": 0.5222,
+      "step": 13936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6272573302022706,
+      "learning_rate": 1.5144689040660206e-05,
+      "loss": 1.2129,
+      "step": 13937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3862090293684135,
+      "learning_rate": 1.5140856360775746e-05,
+      "loss": 0.7285,
+      "step": 13938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3800291721575695,
+      "learning_rate": 1.5137023955267248e-05,
+      "loss": 0.709,
+      "step": 13939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5788851934087096,
+      "learning_rate": 1.5133191824241352e-05,
+      "loss": 0.368,
+      "step": 13940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.398508574088149,
+      "learning_rate": 1.5129359967804705e-05,
+      "loss": 0.6545,
+      "step": 13941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2361272957924188,
+      "learning_rate": 1.512552838606396e-05,
+      "loss": 0.7842,
+      "step": 13942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.728844679458128,
+      "learning_rate": 1.512169707912574e-05,
+      "loss": 0.7939,
+      "step": 13943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9385399027356365,
+      "learning_rate": 1.5117866047096665e-05,
+      "loss": 0.769,
+      "step": 13944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5954718332454334,
+      "learning_rate": 1.5114035290083362e-05,
+      "loss": 0.7834,
+      "step": 13945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.239337560979618,
+      "learning_rate": 1.5110204808192423e-05,
+      "loss": 1.2461,
+      "step": 13946
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.428524217226838,
+      "learning_rate": 1.510637460153047e-05,
+      "loss": 1.4277,
+      "step": 13947
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.438754125891734,
+      "learning_rate": 1.5102544670204074e-05,
+      "loss": 1.4336,
+      "step": 13948
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.18937167068041,
+      "learning_rate": 1.5098715014319837e-05,
+      "loss": 1.0117,
+      "step": 13949
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3240800373618558,
+      "learning_rate": 1.5094885633984324e-05,
+      "loss": 0.9707,
+      "step": 13950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.260373067513544,
+      "learning_rate": 1.5091056529304109e-05,
+      "loss": 0.958,
+      "step": 13951
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8459969752672256,
+      "learning_rate": 1.5087227700385764e-05,
+      "loss": 0.5679,
+      "step": 13952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4082952947373486,
+      "learning_rate": 1.5083399147335836e-05,
+      "loss": 0.4682,
+      "step": 13953
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.618044493697115,
+      "learning_rate": 1.5079570870260873e-05,
+      "loss": 1.1924,
+      "step": 13954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8801947392745895,
+      "learning_rate": 1.5075742869267412e-05,
+      "loss": 0.9922,
+      "step": 13955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.615010098789766,
+      "learning_rate": 1.507191514446199e-05,
+      "loss": 0.5073,
+      "step": 13956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.511228617401004,
+      "learning_rate": 1.5068087695951125e-05,
+      "loss": 0.7529,
+      "step": 13957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7849086999401087,
+      "learning_rate": 1.506426052384134e-05,
+      "loss": 0.9746,
+      "step": 13958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9351897955640993,
+      "learning_rate": 1.5060433628239134e-05,
+      "loss": 1.1387,
+      "step": 13959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.419908928947015,
+      "learning_rate": 1.505660700925101e-05,
+      "loss": 0.791,
+      "step": 13960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0135986160996158,
+      "learning_rate": 1.5052780666983468e-05,
+      "loss": 0.7542,
+      "step": 13961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.334226357912172,
+      "learning_rate": 1.5048954601542997e-05,
+      "loss": 1.0957,
+      "step": 13962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0491408990287283,
+      "learning_rate": 1.5045128813036063e-05,
+      "loss": 0.8635,
+      "step": 13963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.108605064262207,
+      "learning_rate": 1.5041303301569146e-05,
+      "loss": 0.9365,
+      "step": 13964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2934469994757447,
+      "learning_rate": 1.50374780672487e-05,
+      "loss": 0.8525,
+      "step": 13965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.579609682976815,
+      "learning_rate": 1.503365311018119e-05,
+      "loss": 1.1196,
+      "step": 13966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.631182515441743,
+      "learning_rate": 1.5029828430473048e-05,
+      "loss": 1.0566,
+      "step": 13967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9803705134659697,
+      "learning_rate": 1.5026004028230728e-05,
+      "loss": 0.7471,
+      "step": 13968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.905793020765383,
+      "learning_rate": 1.5022179903560646e-05,
+      "loss": 1.1699,
+      "step": 13969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9377447797365144,
+      "learning_rate": 1.501835605656924e-05,
+      "loss": 1.0605,
+      "step": 13970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.613698244519236,
+      "learning_rate": 1.5014532487362928e-05,
+      "loss": 1.0107,
+      "step": 13971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6125022914575617,
+      "learning_rate": 1.5010709196048106e-05,
+      "loss": 1.2363,
+      "step": 13972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0662820001200566,
+      "learning_rate": 1.5006886182731184e-05,
+      "loss": 1.0693,
+      "step": 13973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9102197180191103,
+      "learning_rate": 1.500306344751855e-05,
+      "loss": 0.6772,
+      "step": 13974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7315989091092077,
+      "learning_rate": 1.4999240990516591e-05,
+      "loss": 0.7312,
+      "step": 13975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8545370937734393,
+      "learning_rate": 1.4995418811831679e-05,
+      "loss": 0.8955,
+      "step": 13976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9902305415811954,
+      "learning_rate": 1.49915969115702e-05,
+      "loss": 0.2507,
+      "step": 13977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.077140454438088,
+      "learning_rate": 1.4987775289838488e-05,
+      "loss": 1.1641,
+      "step": 13978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.743954081320737,
+      "learning_rate": 1.498395394674292e-05,
+      "loss": 1.0566,
+      "step": 13979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2253163916413583,
+      "learning_rate": 1.4980132882389835e-05,
+      "loss": 0.7322,
+      "step": 13980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.363419480091661,
+      "learning_rate": 1.497631209688558e-05,
+      "loss": 1.0869,
+      "step": 13981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6296432620780392,
+      "learning_rate": 1.4972491590336473e-05,
+      "loss": 1.0205,
+      "step": 13982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4461387975529236,
+      "learning_rate": 1.4968671362848852e-05,
+      "loss": 0.9727,
+      "step": 13983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.079785706745376,
+      "learning_rate": 1.4964851414529013e-05,
+      "loss": 1.0488,
+      "step": 13984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1264378787935856,
+      "learning_rate": 1.4961031745483281e-05,
+      "loss": 1.3872,
+      "step": 13985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8559675650434428,
+      "learning_rate": 1.4957212355817942e-05,
+      "loss": 0.584,
+      "step": 13986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.7393871001258026,
+      "learning_rate": 1.4953393245639303e-05,
+      "loss": 0.5902,
+      "step": 13987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.505728426076697,
+      "learning_rate": 1.4949574415053624e-05,
+      "loss": 1.3594,
+      "step": 13988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.8678145600539935,
+      "learning_rate": 1.4945755864167208e-05,
+      "loss": 1.5449,
+      "step": 13989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3018308360088753,
+      "learning_rate": 1.4941937593086316e-05,
+      "loss": 1.2637,
+      "step": 13990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.265565753008304,
+      "learning_rate": 1.49381196019172e-05,
+      "loss": 1.0977,
+      "step": 13991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8172282845945467,
+      "learning_rate": 1.4934301890766129e-05,
+      "loss": 0.8691,
+      "step": 13992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.500988704362162,
+      "learning_rate": 1.493048445973933e-05,
+      "loss": 1.7734,
+      "step": 13993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.194045926645283,
+      "learning_rate": 1.4926667308943054e-05,
+      "loss": 1.2051,
+      "step": 13994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.540607815504902,
+      "learning_rate": 1.4922850438483521e-05,
+      "loss": 1.126,
+      "step": 13995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8870292708163927,
+      "learning_rate": 1.4919033848466963e-05,
+      "loss": 0.6331,
+      "step": 13996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234831223227713,
+      "learning_rate": 1.491521753899958e-05,
+      "loss": 0.8945,
+      "step": 13997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4072230469383784,
+      "learning_rate": 1.4911401510187592e-05,
+      "loss": 0.3733,
+      "step": 13998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6439157438451,
+      "learning_rate": 1.4907585762137193e-05,
+      "loss": 0.9629,
+      "step": 13999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.731384484504862,
+      "learning_rate": 1.4903770294954578e-05,
+      "loss": 0.7578,
+      "step": 14000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.445976902570491,
+      "learning_rate": 1.4899955108745922e-05,
+      "loss": 0.8555,
+      "step": 14001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0716947205770464,
+      "learning_rate": 1.4896140203617406e-05,
+      "loss": 0.6738,
+      "step": 14002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.790771179556904,
+      "learning_rate": 1.4892325579675193e-05,
+      "loss": 1.4805,
+      "step": 14003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2311435620648568,
+      "learning_rate": 1.4888511237025446e-05,
+      "loss": 0.834,
+      "step": 14004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.797320490168079,
+      "learning_rate": 1.4884697175774309e-05,
+      "loss": 1.5918,
+      "step": 14005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3689310935656214,
+      "learning_rate": 1.488088339602793e-05,
+      "loss": 1.6543,
+      "step": 14006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.445230185106855,
+      "learning_rate": 1.4877069897892457e-05,
+      "loss": 1.1592,
+      "step": 14007
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5887857753136525,
+      "learning_rate": 1.4873256681474002e-05,
+      "loss": 0.9478,
+      "step": 14008
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.099348360705867,
+      "learning_rate": 1.4869443746878694e-05,
+      "loss": 1.4727,
+      "step": 14009
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.042567885257824,
+      "learning_rate": 1.4865631094212639e-05,
+      "loss": 1.3887,
+      "step": 14010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0383018973588896,
+      "learning_rate": 1.4861818723581949e-05,
+      "loss": 1.0547,
+      "step": 14011
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1766401304862537,
+      "learning_rate": 1.4858006635092714e-05,
+      "loss": 0.999,
+      "step": 14012
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.690153686507041,
+      "learning_rate": 1.4854194828851029e-05,
+      "loss": 1.1934,
+      "step": 14013
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.186423092051741,
+      "learning_rate": 1.4850383304962967e-05,
+      "loss": 1.2734,
+      "step": 14014
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.805716323233035,
+      "learning_rate": 1.4846572063534597e-05,
+      "loss": 1.25,
+      "step": 14015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.345814166418413,
+      "learning_rate": 1.484276110467201e-05,
+      "loss": 1.0059,
+      "step": 14016
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1872897455805496,
+      "learning_rate": 1.4838950428481235e-05,
+      "loss": 1.021,
+      "step": 14017
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5228951169476237,
+      "learning_rate": 1.483514003506834e-05,
+      "loss": 1.0498,
+      "step": 14018
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7339420637419862,
+      "learning_rate": 1.4831329924539356e-05,
+      "loss": 0.9121,
+      "step": 14019
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.714909082954035,
+      "learning_rate": 1.4827520097000319e-05,
+      "loss": 1.123,
+      "step": 14020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.480813790900984,
+      "learning_rate": 1.4823710552557263e-05,
+      "loss": 0.8975,
+      "step": 14021
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1841069746053594,
+      "learning_rate": 1.4819901291316196e-05,
+      "loss": 0.4652,
+      "step": 14022
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9054195078825824,
+      "learning_rate": 1.4816092313383134e-05,
+      "loss": 0.7217,
+      "step": 14023
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.103805230794221,
+      "learning_rate": 1.4812283618864073e-05,
+      "loss": 0.5713,
+      "step": 14024
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.684759383830544,
+      "learning_rate": 1.4808475207865005e-05,
+      "loss": 1.0029,
+      "step": 14025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.72534808464251,
+      "learning_rate": 1.4804667080491935e-05,
+      "loss": 0.8203,
+      "step": 14026
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6780137598711256,
+      "learning_rate": 1.4800859236850823e-05,
+      "loss": 0.78,
+      "step": 14027
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.452548876520361,
+      "learning_rate": 1.4797051677047652e-05,
+      "loss": 0.8643,
+      "step": 14028
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0433721490547114,
+      "learning_rate": 1.4793244401188375e-05,
+      "loss": 1.127,
+      "step": 14029
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8312540256078433,
+      "learning_rate": 1.4789437409378954e-05,
+      "loss": 1.04,
+      "step": 14030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.724243978291281,
+      "learning_rate": 1.4785630701725331e-05,
+      "loss": 0.698,
+      "step": 14031
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.123247632542625,
+      "learning_rate": 1.478182427833345e-05,
+      "loss": 1.5103,
+      "step": 14032
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.821552598007566,
+      "learning_rate": 1.4778018139309236e-05,
+      "loss": 0.9062,
+      "step": 14033
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.308409811008501,
+      "learning_rate": 1.477421228475861e-05,
+      "loss": 1.4746,
+      "step": 14034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.071618343586525,
+      "learning_rate": 1.4770406714787504e-05,
+      "loss": 1.1611,
+      "step": 14035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8050157221345247,
+      "learning_rate": 1.4766601429501808e-05,
+      "loss": 0.7615,
+      "step": 14036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.6112437452572195,
+      "learning_rate": 1.4762796429007435e-05,
+      "loss": 0.9139,
+      "step": 14037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.452755869472565,
+      "learning_rate": 1.4758991713410263e-05,
+      "loss": 0.551,
+      "step": 14038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.77279057298183,
+      "learning_rate": 1.4755187282816185e-05,
+      "loss": 1.0576,
+      "step": 14039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.291858202703926,
+      "learning_rate": 1.4751383137331079e-05,
+      "loss": 1.3594,
+      "step": 14040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0516156216874455,
+      "learning_rate": 1.4747579277060803e-05,
+      "loss": 1.0781,
+      "step": 14041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.1700167857319395,
+      "learning_rate": 1.4743775702111223e-05,
+      "loss": 1.1172,
+      "step": 14042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.242608270941473,
+      "learning_rate": 1.4739972412588183e-05,
+      "loss": 0.8398,
+      "step": 14043
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8801103251125255,
+      "learning_rate": 1.4736169408597538e-05,
+      "loss": 0.7756,
+      "step": 14044
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4367941788143628,
+      "learning_rate": 1.4732366690245128e-05,
+      "loss": 1.2188,
+      "step": 14045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3864839878532154,
+      "learning_rate": 1.4728564257636765e-05,
+      "loss": 1.3301,
+      "step": 14046
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1085465676455164,
+      "learning_rate": 1.4724762110878283e-05,
+      "loss": 0.9209,
+      "step": 14047
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.332027507536143,
+      "learning_rate": 1.4720960250075482e-05,
+      "loss": 0.7544,
+      "step": 14048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.995251712570293,
+      "learning_rate": 1.471715867533418e-05,
+      "loss": 1.4727,
+      "step": 14049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7009544189777877,
+      "learning_rate": 1.471335738676016e-05,
+      "loss": 0.5876,
+      "step": 14050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6145878919120134,
+      "learning_rate": 1.4709556384459222e-05,
+      "loss": 0.8372,
+      "step": 14051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.333780166439181,
+      "learning_rate": 1.470575566853713e-05,
+      "loss": 0.7734,
+      "step": 14052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.452924683768734,
+      "learning_rate": 1.4701955239099672e-05,
+      "loss": 0.6304,
+      "step": 14053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.418046644986722,
+      "learning_rate": 1.469815509625261e-05,
+      "loss": 1.0566,
+      "step": 14054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8861904460963235,
+      "learning_rate": 1.4694355240101695e-05,
+      "loss": 1.041,
+      "step": 14055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4978153219866885,
+      "learning_rate": 1.4690555670752684e-05,
+      "loss": 1.0029,
+      "step": 14056
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.305194361039711,
+      "learning_rate": 1.4686756388311307e-05,
+      "loss": 0.8418,
+      "step": 14057
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1582025210651867,
+      "learning_rate": 1.4682957392883304e-05,
+      "loss": 1.2773,
+      "step": 14058
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4954469227757317,
+      "learning_rate": 1.4679158684574393e-05,
+      "loss": 1.6348,
+      "step": 14059
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9685134140034328,
+      "learning_rate": 1.4675360263490295e-05,
+      "loss": 1.0566,
+      "step": 14060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.128496659100388,
+      "learning_rate": 1.4671562129736721e-05,
+      "loss": 0.6675,
+      "step": 14061
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.125208594566408,
+      "learning_rate": 1.4667764283419357e-05,
+      "loss": 1.6035,
+      "step": 14062
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1793137096541813,
+      "learning_rate": 1.4663966724643912e-05,
+      "loss": 0.8252,
+      "step": 14063
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.506631353193806,
+      "learning_rate": 1.4660169453516072e-05,
+      "loss": 1.2256,
+      "step": 14064
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8147557007938304,
+      "learning_rate": 1.4656372470141505e-05,
+      "loss": 0.873,
+      "step": 14065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.80893294287466,
+      "learning_rate": 1.4652575774625882e-05,
+      "loss": 0.9385,
+      "step": 14066
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.800998227109011,
+      "learning_rate": 1.4648779367074859e-05,
+      "loss": 1.0835,
+      "step": 14067
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4155520685988248,
+      "learning_rate": 1.4644983247594096e-05,
+      "loss": 1.1924,
+      "step": 14068
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.421077790589349,
+      "learning_rate": 1.4641187416289228e-05,
+      "loss": 1.0713,
+      "step": 14069
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.559561649731382,
+      "learning_rate": 1.4637391873265905e-05,
+      "loss": 1.3105,
+      "step": 14070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0642447749937185,
+      "learning_rate": 1.4633596618629736e-05,
+      "loss": 1.2871,
+      "step": 14071
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7199638935114046,
+      "learning_rate": 1.4629801652486358e-05,
+      "loss": 1.0928,
+      "step": 14072
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7592754093972327,
+      "learning_rate": 1.4626006974941383e-05,
+      "loss": 1.2324,
+      "step": 14073
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2640108870052438,
+      "learning_rate": 1.4622212586100406e-05,
+      "loss": 1.3066,
+      "step": 14074
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.106808825838989,
+      "learning_rate": 1.4618418486069033e-05,
+      "loss": 1.1641,
+      "step": 14075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.17830436329256,
+      "learning_rate": 1.4614624674952842e-05,
+      "loss": 1.459,
+      "step": 14076
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3818830835009552,
+      "learning_rate": 1.4610831152857423e-05,
+      "loss": 1.2637,
+      "step": 14077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.152062393087183,
+      "learning_rate": 1.4607037919888339e-05,
+      "loss": 0.5381,
+      "step": 14078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9572418420098465,
+      "learning_rate": 1.460324497615116e-05,
+      "loss": 0.9551,
+      "step": 14079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5463553056397714,
+      "learning_rate": 1.4599452321751433e-05,
+      "loss": 1.1416,
+      "step": 14080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.0065287880297955,
+      "learning_rate": 1.4595659956794722e-05,
+      "loss": 0.918,
+      "step": 14081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1027294450135523,
+      "learning_rate": 1.4591867881386562e-05,
+      "loss": 0.345,
+      "step": 14082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.397977381434268,
+      "learning_rate": 1.458807609563248e-05,
+      "loss": 1.1875,
+      "step": 14083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4680095479647592,
+      "learning_rate": 1.4584284599638005e-05,
+      "loss": 0.7441,
+      "step": 14084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.816452781037905,
+      "learning_rate": 1.4580493393508637e-05,
+      "loss": 0.6523,
+      "step": 14085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.032424400836538,
+      "learning_rate": 1.4576702477349908e-05,
+      "loss": 1.5371,
+      "step": 14086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5983991206047845,
+      "learning_rate": 1.4572911851267306e-05,
+      "loss": 1.0669,
+      "step": 14087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.178838858617146,
+      "learning_rate": 1.4569121515366319e-05,
+      "loss": 0.7069,
+      "step": 14088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8011159444153,
+      "learning_rate": 1.4565331469752427e-05,
+      "loss": 1.1025,
+      "step": 14089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.894933410118725,
+      "learning_rate": 1.4561541714531112e-05,
+      "loss": 1.1138,
+      "step": 14090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7395134491122322,
+      "learning_rate": 1.455775224980785e-05,
+      "loss": 1.3809,
+      "step": 14091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.4840976150196505,
+      "learning_rate": 1.4553963075688093e-05,
+      "loss": 1.1367,
+      "step": 14092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.639111204524455,
+      "learning_rate": 1.4550174192277288e-05,
+      "loss": 1.4277,
+      "step": 14093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.180611144678123,
+      "learning_rate": 1.4546385599680876e-05,
+      "loss": 0.8896,
+      "step": 14094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.534196433892163,
+      "learning_rate": 1.4542597298004304e-05,
+      "loss": 1.0176,
+      "step": 14095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.413194617645721,
+      "learning_rate": 1.4538809287352995e-05,
+      "loss": 1.3438,
+      "step": 14096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.476889331327968,
+      "learning_rate": 1.453502156783236e-05,
+      "loss": 0.7529,
+      "step": 14097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.97107016321555,
+      "learning_rate": 1.4531234139547806e-05,
+      "loss": 0.8838,
+      "step": 14098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5134131143523617,
+      "learning_rate": 1.4527447002604746e-05,
+      "loss": 0.9121,
+      "step": 14099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.749781426499963,
+      "learning_rate": 1.4523660157108583e-05,
+      "loss": 1.0352,
+      "step": 14100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.130819553271437,
+      "learning_rate": 1.4519873603164692e-05,
+      "loss": 0.9473,
+      "step": 14101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1459287582644326,
+      "learning_rate": 1.451608734087844e-05,
+      "loss": 0.7744,
+      "step": 14102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9042762794359018,
+      "learning_rate": 1.451230137035522e-05,
+      "loss": 0.7028,
+      "step": 14103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8708740615093042,
+      "learning_rate": 1.4508515691700386e-05,
+      "loss": 0.6925,
+      "step": 14104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4231553169815725,
+      "learning_rate": 1.4504730305019287e-05,
+      "loss": 1.0762,
+      "step": 14105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.019361842731099,
+      "learning_rate": 1.450094521041726e-05,
+      "loss": 1.2295,
+      "step": 14106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.104116810658808,
+      "learning_rate": 1.4497160407999667e-05,
+      "loss": 0.9238,
+      "step": 14107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.43246417701934,
+      "learning_rate": 1.4493375897871811e-05,
+      "loss": 1.3926,
+      "step": 14108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.901190326677977,
+      "learning_rate": 1.4489591680139037e-05,
+      "loss": 1.5049,
+      "step": 14109
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.268270381957425,
+      "learning_rate": 1.4485807754906649e-05,
+      "loss": 0.7712,
+      "step": 14110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8388972535065085,
+      "learning_rate": 1.4482024122279941e-05,
+      "loss": 0.5186,
+      "step": 14111
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1562571870988125,
+      "learning_rate": 1.4478240782364227e-05,
+      "loss": 0.8545,
+      "step": 14112
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.0793393057176015,
+      "learning_rate": 1.447445773526479e-05,
+      "loss": 0.8916,
+      "step": 14113
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5919515970326032,
+      "learning_rate": 1.4470674981086913e-05,
+      "loss": 1.5586,
+      "step": 14114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.678649277699783,
+      "learning_rate": 1.4466892519935852e-05,
+      "loss": 1.5781,
+      "step": 14115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6046711445771464,
+      "learning_rate": 1.4463110351916892e-05,
+      "loss": 0.9644,
+      "step": 14116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8246512813885887,
+      "learning_rate": 1.4459328477135276e-05,
+      "loss": 1.2656,
+      "step": 14117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.62645553544976,
+      "learning_rate": 1.4455546895696264e-05,
+      "loss": 0.9624,
+      "step": 14118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.271499174533957,
+      "learning_rate": 1.4451765607705078e-05,
+      "loss": 0.7505,
+      "step": 14119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.329846433930485,
+      "learning_rate": 1.4447984613266974e-05,
+      "loss": 0.9531,
+      "step": 14120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.562696123759017,
+      "learning_rate": 1.4444203912487162e-05,
+      "loss": 0.752,
+      "step": 14121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.115782409299384,
+      "learning_rate": 1.4440423505470856e-05,
+      "loss": 0.6345,
+      "step": 14122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5584954395771664,
+      "learning_rate": 1.4436643392323256e-05,
+      "loss": 1.1543,
+      "step": 14123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.168425270959632,
+      "learning_rate": 1.443286357314958e-05,
+      "loss": 1.2207,
+      "step": 14124
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.304328917038977,
+      "learning_rate": 1.4429084048055008e-05,
+      "loss": 1.5469,
+      "step": 14125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7960958407816032,
+      "learning_rate": 1.4425304817144726e-05,
+      "loss": 0.7002,
+      "step": 14126
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.473105632329207,
+      "learning_rate": 1.4421525880523896e-05,
+      "loss": 1.332,
+      "step": 14127
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.186216795436342,
+      "learning_rate": 1.4417747238297691e-05,
+      "loss": 0.6108,
+      "step": 14128
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.714518004709913,
+      "learning_rate": 1.4413968890571283e-05,
+      "loss": 1.0439,
+      "step": 14129
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.345627401560594,
+      "learning_rate": 1.4410190837449817e-05,
+      "loss": 1.0352,
+      "step": 14130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.25984823574469,
+      "learning_rate": 1.4406413079038422e-05,
+      "loss": 1.3086,
+      "step": 14131
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1234331525318813,
+      "learning_rate": 1.4402635615442231e-05,
+      "loss": 0.9766,
+      "step": 14132
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.841425689890882,
+      "learning_rate": 1.4398858446766387e-05,
+      "loss": 1.4336,
+      "step": 14133
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4871222702849596,
+      "learning_rate": 1.4395081573115992e-05,
+      "loss": 0.7451,
+      "step": 14134
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.7057900988332,
+      "learning_rate": 1.4391304994596166e-05,
+      "loss": 1.2134,
+      "step": 14135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.171357127408975,
+      "learning_rate": 1.4387528711311981e-05,
+      "loss": 1.1006,
+      "step": 14136
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.366436729096335,
+      "learning_rate": 1.4383752723368571e-05,
+      "loss": 1.2493,
+      "step": 14137
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4833357420065796,
+      "learning_rate": 1.4379977030871e-05,
+      "loss": 1.2871,
+      "step": 14138
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.114738258236321,
+      "learning_rate": 1.4376201633924346e-05,
+      "loss": 0.9176,
+      "step": 14139
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4765295754539545,
+      "learning_rate": 1.4372426532633664e-05,
+      "loss": 0.7866,
+      "step": 14140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.467687257349571,
+      "learning_rate": 1.4368651727104037e-05,
+      "loss": 1.0166,
+      "step": 14141
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.785955569400982,
+      "learning_rate": 1.4364877217440506e-05,
+      "loss": 1.2559,
+      "step": 14142
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.4393050246466395,
+      "learning_rate": 1.4361103003748111e-05,
+      "loss": 1.4395,
+      "step": 14143
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2906717816946296,
+      "learning_rate": 1.4357329086131877e-05,
+      "loss": 1.1812,
+      "step": 14144
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.187361529893407,
+      "learning_rate": 1.4353555464696843e-05,
+      "loss": 0.8057,
+      "step": 14145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.453685967029629,
+      "learning_rate": 1.4349782139548038e-05,
+      "loss": 1.5293,
+      "step": 14146
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.262442412091363,
+      "learning_rate": 1.4346009110790459e-05,
+      "loss": 1.3945,
+      "step": 14147
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.519596145827547,
+      "learning_rate": 1.4342236378529105e-05,
+      "loss": 1.2959,
+      "step": 14148
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.609707878245359,
+      "learning_rate": 1.4338463942868968e-05,
+      "loss": 1.5645,
+      "step": 14149
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.06880127073988,
+      "learning_rate": 1.433469180391505e-05,
+      "loss": 1.0396,
+      "step": 14150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3744541344003363,
+      "learning_rate": 1.4330919961772316e-05,
+      "loss": 1.0732,
+      "step": 14151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2648593424807877,
+      "learning_rate": 1.4327148416545735e-05,
+      "loss": 1.1309,
+      "step": 14152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6208213971649044,
+      "learning_rate": 1.4323377168340261e-05,
+      "loss": 1.2227,
+      "step": 14153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.750206506084741,
+      "learning_rate": 1.431960621726085e-05,
+      "loss": 1.1343,
+      "step": 14154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7376684246580174,
+      "learning_rate": 1.4315835563412464e-05,
+      "loss": 0.5439,
+      "step": 14155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7270734902512863,
+      "learning_rate": 1.4312065206900022e-05,
+      "loss": 1.4082,
+      "step": 14156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3679889795137115,
+      "learning_rate": 1.4308295147828448e-05,
+      "loss": 0.8857,
+      "step": 14157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.913584979411271,
+      "learning_rate": 1.4304525386302676e-05,
+      "loss": 0.9932,
+      "step": 14158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7177267510038168,
+      "learning_rate": 1.4300755922427606e-05,
+      "loss": 1.1914,
+      "step": 14159
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.703356332055007,
+      "learning_rate": 1.4296986756308145e-05,
+      "loss": 1.3867,
+      "step": 14160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.698894435634442,
+      "learning_rate": 1.4293217888049173e-05,
+      "loss": 1.2949,
+      "step": 14161
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9332255822848676,
+      "learning_rate": 1.4289449317755599e-05,
+      "loss": 1.2217,
+      "step": 14162
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6731250988692667,
+      "learning_rate": 1.4285681045532292e-05,
+      "loss": 1.5537,
+      "step": 14163
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.751642407248505,
+      "learning_rate": 1.4281913071484109e-05,
+      "loss": 1.5918,
+      "step": 14164
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.40522820782086,
+      "learning_rate": 1.427814539571593e-05,
+      "loss": 0.9863,
+      "step": 14165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.215557997054613,
+      "learning_rate": 1.4274378018332597e-05,
+      "loss": 0.5952,
+      "step": 14166
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9464717779253125,
+      "learning_rate": 1.427061093943896e-05,
+      "loss": 0.7773,
+      "step": 14167
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6860017836709502,
+      "learning_rate": 1.4266844159139853e-05,
+      "loss": 0.8525,
+      "step": 14168
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.564112966037108,
+      "learning_rate": 1.4263077677540104e-05,
+      "loss": 0.8843,
+      "step": 14169
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.277163913234885,
+      "learning_rate": 1.4259311494744525e-05,
+      "loss": 0.7236,
+      "step": 14170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6544681536969184,
+      "learning_rate": 1.4255545610857943e-05,
+      "loss": 0.6143,
+      "step": 14171
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5350644604916175,
+      "learning_rate": 1.4251780025985153e-05,
+      "loss": 1.3047,
+      "step": 14172
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6567489414006222,
+      "learning_rate": 1.4248014740230939e-05,
+      "loss": 0.8374,
+      "step": 14173
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.143217772068376,
+      "learning_rate": 1.4244249753700107e-05,
+      "loss": 0.7793,
+      "step": 14174
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8643120839556633,
+      "learning_rate": 1.4240485066497417e-05,
+      "loss": 0.8115,
+      "step": 14175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.658928842720384,
+      "learning_rate": 1.4236720678727655e-05,
+      "loss": 1.373,
+      "step": 14176
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.347822098985023,
+      "learning_rate": 1.4232956590495577e-05,
+      "loss": 0.8781,
+      "step": 14177
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4238505519963653,
+      "learning_rate": 1.4229192801905921e-05,
+      "loss": 1.1221,
+      "step": 14178
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.728196749045101,
+      "learning_rate": 1.4225429313063459e-05,
+      "loss": 0.5757,
+      "step": 14179
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.043309399558387,
+      "learning_rate": 1.422166612407291e-05,
+      "loss": 1.2246,
+      "step": 14180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2049818657943763,
+      "learning_rate": 1.4217903235039007e-05,
+      "loss": 0.8857,
+      "step": 14181
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.018797064497565,
+      "learning_rate": 1.4214140646066453e-05,
+      "loss": 0.719,
+      "step": 14182
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4385188586844073,
+      "learning_rate": 1.421037835725998e-05,
+      "loss": 0.373,
+      "step": 14183
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.392611559871641,
+      "learning_rate": 1.420661636872429e-05,
+      "loss": 1.457,
+      "step": 14184
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.249982721630204,
+      "learning_rate": 1.4202854680564076e-05,
+      "loss": 0.9346,
+      "step": 14185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5406412472735123,
+      "learning_rate": 1.4199093292884019e-05,
+      "loss": 0.9863,
+      "step": 14186
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.137179123137898,
+      "learning_rate": 1.419533220578879e-05,
+      "loss": 1.5293,
+      "step": 14187
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1841880196883103,
+      "learning_rate": 1.4191571419383077e-05,
+      "loss": 0.9634,
+      "step": 14188
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.039096187508519,
+      "learning_rate": 1.4187810933771529e-05,
+      "loss": 1.0527,
+      "step": 14189
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.503585540339952,
+      "learning_rate": 1.4184050749058803e-05,
+      "loss": 1.6108,
+      "step": 14190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.593040774864764,
+      "learning_rate": 1.4180290865349526e-05,
+      "loss": 0.7764,
+      "step": 14191
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.924725001059406,
+      "learning_rate": 1.4176531282748356e-05,
+      "loss": 1.1963,
+      "step": 14192
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7169083178286555,
+      "learning_rate": 1.4172772001359924e-05,
+      "loss": 0.958,
+      "step": 14193
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.846966549336602,
+      "learning_rate": 1.4169013021288835e-05,
+      "loss": 1.4512,
+      "step": 14194
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.849131492639703,
+      "learning_rate": 1.4165254342639708e-05,
+      "loss": 1.1641,
+      "step": 14195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.549964294931603,
+      "learning_rate": 1.416149596551713e-05,
+      "loss": 0.7788,
+      "step": 14196
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.836404001939497,
+      "learning_rate": 1.4157737890025718e-05,
+      "loss": 1.2988,
+      "step": 14197
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.268186303981047,
+      "learning_rate": 1.415398011627005e-05,
+      "loss": 1.2129,
+      "step": 14198
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.113747688391776,
+      "learning_rate": 1.4150222644354682e-05,
+      "loss": 1.7705,
+      "step": 14199
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.826526739737573,
+      "learning_rate": 1.4146465474384219e-05,
+      "loss": 1.2207,
+      "step": 14200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7331907487136275,
+      "learning_rate": 1.414270860646319e-05,
+      "loss": 1.3516,
+      "step": 14201
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5958785677849816,
+      "learning_rate": 1.4138952040696169e-05,
+      "loss": 0.7539,
+      "step": 14202
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.000009775149799,
+      "learning_rate": 1.4135195777187692e-05,
+      "loss": 1.4785,
+      "step": 14203
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.070175666485992,
+      "learning_rate": 1.4131439816042283e-05,
+      "loss": 1.2207,
+      "step": 14204
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3320499591939523,
+      "learning_rate": 1.412768415736449e-05,
+      "loss": 1.082,
+      "step": 14205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.868142611380773,
+      "learning_rate": 1.4123928801258821e-05,
+      "loss": 1.2632,
+      "step": 14206
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2828692607363634,
+      "learning_rate": 1.4120173747829785e-05,
+      "loss": 1.4336,
+      "step": 14207
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.420814245235019,
+      "learning_rate": 1.4116418997181873e-05,
+      "loss": 1.0615,
+      "step": 14208
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9538369909932998,
+      "learning_rate": 1.4112664549419601e-05,
+      "loss": 0.8579,
+      "step": 14209
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9716386946658018,
+      "learning_rate": 1.4108910404647434e-05,
+      "loss": 0.4716,
+      "step": 14210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.4547332536534725,
+      "learning_rate": 1.4105156562969869e-05,
+      "loss": 0.9493,
+      "step": 14211
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1446393675609547,
+      "learning_rate": 1.4101403024491355e-05,
+      "loss": 0.7954,
+      "step": 14212
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4811357217219077,
+      "learning_rate": 1.4097649789316352e-05,
+      "loss": 1.082,
+      "step": 14213
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.62463514193376,
+      "learning_rate": 1.4093896857549327e-05,
+      "loss": 0.6445,
+      "step": 14214
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.500627340707504,
+      "learning_rate": 1.409014422929471e-05,
+      "loss": 1.9219,
+      "step": 14215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.697616646512652,
+      "learning_rate": 1.4086391904656932e-05,
+      "loss": 1.7734,
+      "step": 14216
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.133918341780628,
+      "learning_rate": 1.4082639883740431e-05,
+      "loss": 1.1641,
+      "step": 14217
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.970503921400897,
+      "learning_rate": 1.4078888166649618e-05,
+      "loss": 1.1855,
+      "step": 14218
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.140831935764514,
+      "learning_rate": 1.4075136753488894e-05,
+      "loss": 1.1064,
+      "step": 14219
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7037530070308873,
+      "learning_rate": 1.4071385644362672e-05,
+      "loss": 0.6929,
+      "step": 14220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.555669963174862,
+      "learning_rate": 1.4067634839375333e-05,
+      "loss": 1.1719,
+      "step": 14221
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.953497484724331,
+      "learning_rate": 1.4063884338631272e-05,
+      "loss": 0.8423,
+      "step": 14222
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8339389452983474,
+      "learning_rate": 1.406013414223486e-05,
+      "loss": 1.7969,
+      "step": 14223
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.76216390635772,
+      "learning_rate": 1.4056384250290461e-05,
+      "loss": 0.707,
+      "step": 14224
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.7040613199246595,
+      "learning_rate": 1.4052634662902426e-05,
+      "loss": 0.8994,
+      "step": 14225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8339138484633937,
+      "learning_rate": 1.4048885380175115e-05,
+      "loss": 0.7412,
+      "step": 14226
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2346983989011835,
+      "learning_rate": 1.4045136402212872e-05,
+      "loss": 1.3291,
+      "step": 14227
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.431410102181747,
+      "learning_rate": 1.404138772912002e-05,
+      "loss": 1.1582,
+      "step": 14228
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.323414799626782,
+      "learning_rate": 1.4037639361000876e-05,
+      "loss": 1.2402,
+      "step": 14229
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.5175608145204444,
+      "learning_rate": 1.4033891297959767e-05,
+      "loss": 0.6973,
+      "step": 14230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.483931111920842,
+      "learning_rate": 1.4030143540101007e-05,
+      "loss": 0.894,
+      "step": 14231
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.513875127266015,
+      "learning_rate": 1.4026396087528887e-05,
+      "loss": 1.1875,
+      "step": 14232
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.46744270961618,
+      "learning_rate": 1.4022648940347699e-05,
+      "loss": 0.8877,
+      "step": 14233
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1374199298505046,
+      "learning_rate": 1.401890209866171e-05,
+      "loss": 1.0986,
+      "step": 14234
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2673098949418558,
+      "learning_rate": 1.4015155562575215e-05,
+      "loss": 1.3701,
+      "step": 14235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.5931300568873965,
+      "learning_rate": 1.4011409332192472e-05,
+      "loss": 1.0059,
+      "step": 14236
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7345772015349206,
+      "learning_rate": 1.400766340761772e-05,
+      "loss": 0.9004,
+      "step": 14237
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.737674583096887,
+      "learning_rate": 1.400391778895522e-05,
+      "loss": 1.1758,
+      "step": 14238
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.428575463652264,
+      "learning_rate": 1.4000172476309224e-05,
+      "loss": 0.8779,
+      "step": 14239
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.884903147111336,
+      "learning_rate": 1.3996427469783948e-05,
+      "loss": 1.0828,
+      "step": 14240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.887541748018963,
+      "learning_rate": 1.3992682769483616e-05,
+      "loss": 1.3926,
+      "step": 14241
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.036309667331141,
+      "learning_rate": 1.3988938375512433e-05,
+      "loss": 1.1699,
+      "step": 14242
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.095761742661129,
+      "learning_rate": 1.3985194287974624e-05,
+      "loss": 1.188,
+      "step": 14243
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4310124994293205,
+      "learning_rate": 1.3981450506974369e-05,
+      "loss": 0.5247,
+      "step": 14244
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4645742519939033,
+      "learning_rate": 1.3977707032615867e-05,
+      "loss": 0.9189,
+      "step": 14245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6758245775707583,
+      "learning_rate": 1.3973963865003276e-05,
+      "loss": 0.6772,
+      "step": 14246
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.488017548821187,
+      "learning_rate": 1.3970221004240784e-05,
+      "loss": 1.1914,
+      "step": 14247
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.287033796433198,
+      "learning_rate": 1.3966478450432557e-05,
+      "loss": 0.6162,
+      "step": 14248
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0745325182852845,
+      "learning_rate": 1.3962736203682746e-05,
+      "loss": 1.1074,
+      "step": 14249
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8777465930791197,
+      "learning_rate": 1.3958994264095496e-05,
+      "loss": 1.0078,
+      "step": 14250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.564607730339351,
+      "learning_rate": 1.3955252631774928e-05,
+      "loss": 0.8164,
+      "step": 14251
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.374553889204871,
+      "learning_rate": 1.3951511306825193e-05,
+      "loss": 0.8438,
+      "step": 14252
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.402080538174557,
+      "learning_rate": 1.3947770289350404e-05,
+      "loss": 0.8975,
+      "step": 14253
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7476345833401778,
+      "learning_rate": 1.3944029579454665e-05,
+      "loss": 0.6914,
+      "step": 14254
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.012601652495515,
+      "learning_rate": 1.3940289177242072e-05,
+      "loss": 1.3887,
+      "step": 14255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.746375690543347,
+      "learning_rate": 1.393654908281673e-05,
+      "loss": 1.1089,
+      "step": 14256
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7353995605576333,
+      "learning_rate": 1.393280929628273e-05,
+      "loss": 0.771,
+      "step": 14257
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.657878545647998,
+      "learning_rate": 1.3929069817744144e-05,
+      "loss": 0.9336,
+      "step": 14258
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.392639248800386,
+      "learning_rate": 1.3925330647305029e-05,
+      "loss": 0.6357,
+      "step": 14259
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.95047224470117,
+      "learning_rate": 1.3921591785069463e-05,
+      "loss": 1.5938,
+      "step": 14260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1536646284384924,
+      "learning_rate": 1.3917853231141487e-05,
+      "loss": 1.2266,
+      "step": 14261
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.612570565032865,
+      "learning_rate": 1.3914114985625143e-05,
+      "loss": 1.3301,
+      "step": 14262
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.13439268046215,
+      "learning_rate": 1.391037704862446e-05,
+      "loss": 0.4655,
+      "step": 14263
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.250894561121619,
+      "learning_rate": 1.3906639420243476e-05,
+      "loss": 1.248,
+      "step": 14264
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3207982643349756,
+      "learning_rate": 1.3902902100586202e-05,
+      "loss": 0.8535,
+      "step": 14265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4239072295482442,
+      "learning_rate": 1.3899165089756632e-05,
+      "loss": 0.7881,
+      "step": 14266
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.129793855864201,
+      "learning_rate": 1.3895428387858791e-05,
+      "loss": 1.2656,
+      "step": 14267
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.053726823691961,
+      "learning_rate": 1.3891691994996647e-05,
+      "loss": 1.417,
+      "step": 14268
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4865841555147252,
+      "learning_rate": 1.3887955911274203e-05,
+      "loss": 0.6421,
+      "step": 14269
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.339521127917248,
+      "learning_rate": 1.3884220136795423e-05,
+      "loss": 0.5042,
+      "step": 14270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9622929968058058,
+      "learning_rate": 1.3880484671664267e-05,
+      "loss": 1.0059,
+      "step": 14271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.860141693439843,
+      "learning_rate": 1.3876749515984692e-05,
+      "loss": 1.0586,
+      "step": 14272
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6407652146170983,
+      "learning_rate": 1.3873014669860657e-05,
+      "loss": 0.9004,
+      "step": 14273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5723468185505673,
+      "learning_rate": 1.386928013339609e-05,
+      "loss": 1.0957,
+      "step": 14274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.817309357794612,
+      "learning_rate": 1.3865545906694922e-05,
+      "loss": 0.8457,
+      "step": 14275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.797619028542714,
+      "learning_rate": 1.3861811989861073e-05,
+      "loss": 1.0322,
+      "step": 14276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.447595182911758,
+      "learning_rate": 1.3858078382998473e-05,
+      "loss": 0.4116,
+      "step": 14277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0027335083410693,
+      "learning_rate": 1.385434508621102e-05,
+      "loss": 1.4492,
+      "step": 14278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7375218185879127,
+      "learning_rate": 1.3850612099602603e-05,
+      "loss": 0.7993,
+      "step": 14279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3657594560066593,
+      "learning_rate": 1.3846879423277104e-05,
+      "loss": 1.041,
+      "step": 14280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.908647457640573,
+      "learning_rate": 1.3843147057338418e-05,
+      "loss": 0.6938,
+      "step": 14281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6551538337351266,
+      "learning_rate": 1.3839415001890411e-05,
+      "loss": 1.1846,
+      "step": 14282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.112262596072583,
+      "learning_rate": 1.3835683257036936e-05,
+      "loss": 1.0176,
+      "step": 14283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5715478141496755,
+      "learning_rate": 1.3831951822881842e-05,
+      "loss": 1.0459,
+      "step": 14284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6935788497388655,
+      "learning_rate": 1.3828220699528982e-05,
+      "loss": 0.6284,
+      "step": 14285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5774876587033004,
+      "learning_rate": 1.3824489887082204e-05,
+      "loss": 1.373,
+      "step": 14286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9776796662440344,
+      "learning_rate": 1.3820759385645323e-05,
+      "loss": 1.3066,
+      "step": 14287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2353969984324005,
+      "learning_rate": 1.3817029195322156e-05,
+      "loss": 1.2129,
+      "step": 14288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.141650211948855,
+      "learning_rate": 1.3813299316216504e-05,
+      "loss": 0.9182,
+      "step": 14289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2679428346147357,
+      "learning_rate": 1.3809569748432189e-05,
+      "loss": 0.9888,
+      "step": 14290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8179586208871097,
+      "learning_rate": 1.3805840492072991e-05,
+      "loss": 0.5073,
+      "step": 14291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7905584859236683,
+      "learning_rate": 1.3802111547242696e-05,
+      "loss": 0.9678,
+      "step": 14292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.147726711256606,
+      "learning_rate": 1.3798382914045058e-05,
+      "loss": 0.8857,
+      "step": 14293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4690955197910127,
+      "learning_rate": 1.3794654592583883e-05,
+      "loss": 0.6562,
+      "step": 14294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.596557109302412,
+      "learning_rate": 1.379092658296291e-05,
+      "loss": 1.2734,
+      "step": 14295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.566257967144407,
+      "learning_rate": 1.3787198885285885e-05,
+      "loss": 1.2578,
+      "step": 14296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4654325825665637,
+      "learning_rate": 1.3783471499656541e-05,
+      "loss": 0.9111,
+      "step": 14297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5465449956052875,
+      "learning_rate": 1.3779744426178629e-05,
+      "loss": 0.6487,
+      "step": 14298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.912289642431508,
+      "learning_rate": 1.3776017664955863e-05,
+      "loss": 0.7346,
+      "step": 14299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.446095594114853,
+      "learning_rate": 1.3772291216091954e-05,
+      "loss": 1.1377,
+      "step": 14300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.094545909060534,
+      "learning_rate": 1.37685650796906e-05,
+      "loss": 1.5928,
+      "step": 14301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.718726406503412,
+      "learning_rate": 1.3764839255855518e-05,
+      "loss": 1.25,
+      "step": 14302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4528794864329106,
+      "learning_rate": 1.3761113744690376e-05,
+      "loss": 0.8596,
+      "step": 14303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.897536083872292,
+      "learning_rate": 1.3757388546298872e-05,
+      "loss": 1.0371,
+      "step": 14304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.412127520058853,
+      "learning_rate": 1.3753663660784672e-05,
+      "loss": 0.9541,
+      "step": 14305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4918901511239526,
+      "learning_rate": 1.3749939088251418e-05,
+      "loss": 1.7266,
+      "step": 14306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.71147231170832,
+      "learning_rate": 1.3746214828802791e-05,
+      "loss": 0.8643,
+      "step": 14307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1967141234964216,
+      "learning_rate": 1.3742490882542425e-05,
+      "loss": 1.2656,
+      "step": 14308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.358332624761073,
+      "learning_rate": 1.3738767249573956e-05,
+      "loss": 0.9727,
+      "step": 14309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.295509763860158,
+      "learning_rate": 1.3735043930001e-05,
+      "loss": 1.7402,
+      "step": 14310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.313244574014839,
+      "learning_rate": 1.3731320923927193e-05,
+      "loss": 1.1943,
+      "step": 14311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.051003105324241,
+      "learning_rate": 1.3727598231456129e-05,
+      "loss": 0.9463,
+      "step": 14312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6550253096720495,
+      "learning_rate": 1.3723875852691426e-05,
+      "loss": 1.1567,
+      "step": 14313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.272259265738195,
+      "learning_rate": 1.372015378773666e-05,
+      "loss": 1.1455,
+      "step": 14314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.6524937987452795,
+      "learning_rate": 1.3716432036695428e-05,
+      "loss": 1.2637,
+      "step": 14315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7096482135323914,
+      "learning_rate": 1.3712710599671302e-05,
+      "loss": 1.3379,
+      "step": 14316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4882971555000872,
+      "learning_rate": 1.3708989476767841e-05,
+      "loss": 1.2715,
+      "step": 14317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5325525958821515,
+      "learning_rate": 1.37052686680886e-05,
+      "loss": 1.4844,
+      "step": 14318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.444784542770563,
+      "learning_rate": 1.3701548173737144e-05,
+      "loss": 1.0635,
+      "step": 14319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.176983966105279,
+      "learning_rate": 1.3697827993817003e-05,
+      "loss": 1.0728,
+      "step": 14320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8893129193816807,
+      "learning_rate": 1.36941081284317e-05,
+      "loss": 0.4438,
+      "step": 14321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.688857044063374,
+      "learning_rate": 1.3690388577684768e-05,
+      "loss": 1.791,
+      "step": 14322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2702392944379257,
+      "learning_rate": 1.3686669341679712e-05,
+      "loss": 0.8213,
+      "step": 14323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.4480427455952425,
+      "learning_rate": 1.3682950420520051e-05,
+      "loss": 2.3047,
+      "step": 14324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0834128567174885,
+      "learning_rate": 1.3679231814309274e-05,
+      "loss": 1.2441,
+      "step": 14325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2289994673251257,
+      "learning_rate": 1.3675513523150867e-05,
+      "loss": 1.0771,
+      "step": 14326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.231399689255977,
+      "learning_rate": 1.3671795547148298e-05,
+      "loss": 1.4316,
+      "step": 14327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9147948533603243,
+      "learning_rate": 1.3668077886405055e-05,
+      "loss": 0.8301,
+      "step": 14328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.501722865451837,
+      "learning_rate": 1.3664360541024595e-05,
+      "loss": 1.3398,
+      "step": 14329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5509712650859457,
+      "learning_rate": 1.3660643511110365e-05,
+      "loss": 1.125,
+      "step": 14330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.317809324885962,
+      "learning_rate": 1.3656926796765801e-05,
+      "loss": 1.0811,
+      "step": 14331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5498149729900748,
+      "learning_rate": 1.3653210398094346e-05,
+      "loss": 1.2344,
+      "step": 14332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9829710637372036,
+      "learning_rate": 1.3649494315199434e-05,
+      "loss": 0.8394,
+      "step": 14333
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.174264046693702,
+      "learning_rate": 1.3645778548184479e-05,
+      "loss": 1.4551,
+      "step": 14334
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1616081339083775,
+      "learning_rate": 1.3642063097152868e-05,
+      "loss": 1.1611,
+      "step": 14335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.758513400561692,
+      "learning_rate": 1.3638347962208032e-05,
+      "loss": 1.3457,
+      "step": 14336
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.827587772157202,
+      "learning_rate": 1.3634633143453346e-05,
+      "loss": 0.7412,
+      "step": 14337
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7545225211166686,
+      "learning_rate": 1.3630918640992194e-05,
+      "loss": 1.3008,
+      "step": 14338
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.56587971963079,
+      "learning_rate": 1.3627204454927936e-05,
+      "loss": 0.8711,
+      "step": 14339
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2333713224605662,
+      "learning_rate": 1.3623490585363948e-05,
+      "loss": 0.7358,
+      "step": 14340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.810486662248166,
+      "learning_rate": 1.3619777032403596e-05,
+      "loss": 1.6719,
+      "step": 14341
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.015410537547419,
+      "learning_rate": 1.361606379615022e-05,
+      "loss": 1.4629,
+      "step": 14342
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.438855503445625,
+      "learning_rate": 1.3612350876707148e-05,
+      "loss": 0.5537,
+      "step": 14343
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.999387002347747,
+      "learning_rate": 1.360863827417771e-05,
+      "loss": 1.8516,
+      "step": 14344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2711411247128206,
+      "learning_rate": 1.3604925988665237e-05,
+      "loss": 1.1348,
+      "step": 14345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.149723521755014,
+      "learning_rate": 1.3601214020273037e-05,
+      "loss": 0.9941,
+      "step": 14346
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8792918139878867,
+      "learning_rate": 1.3597502369104409e-05,
+      "loss": 0.9277,
+      "step": 14347
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.917441528571196,
+      "learning_rate": 1.3593791035262643e-05,
+      "loss": 1.1289,
+      "step": 14348
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.025955863850942,
+      "learning_rate": 1.3590080018851026e-05,
+      "loss": 1.4033,
+      "step": 14349
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7126099237956594,
+      "learning_rate": 1.3586369319972845e-05,
+      "loss": 0.9626,
+      "step": 14350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9839678581490783,
+      "learning_rate": 1.358265893873136e-05,
+      "loss": 1.502,
+      "step": 14351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.964629194414704,
+      "learning_rate": 1.357894887522983e-05,
+      "loss": 1.0391,
+      "step": 14352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.823466811584759,
+      "learning_rate": 1.3575239129571492e-05,
+      "loss": 0.9746,
+      "step": 14353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.175420707265855,
+      "learning_rate": 1.357152970185961e-05,
+      "loss": 0.7473,
+      "step": 14354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1276255449913783,
+      "learning_rate": 1.3567820592197402e-05,
+      "loss": 1.1436,
+      "step": 14355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.37822597469204,
+      "learning_rate": 1.3564111800688085e-05,
+      "loss": 1.2578,
+      "step": 14356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9481551722005,
+      "learning_rate": 1.356040332743489e-05,
+      "loss": 1.0352,
+      "step": 14357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.125906021847346,
+      "learning_rate": 1.3556695172541004e-05,
+      "loss": 1.04,
+      "step": 14358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.81445274997598,
+      "learning_rate": 1.3552987336109641e-05,
+      "loss": 0.7041,
+      "step": 14359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.798012346836825,
+      "learning_rate": 1.3549279818243981e-05,
+      "loss": 1.0312,
+      "step": 14360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.142534864512142,
+      "learning_rate": 1.3545572619047192e-05,
+      "loss": 1.5425,
+      "step": 14361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.915282403886731,
+      "learning_rate": 1.3541865738622467e-05,
+      "loss": 1.0762,
+      "step": 14362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5267539891307456,
+      "learning_rate": 1.3538159177072949e-05,
+      "loss": 1.1875,
+      "step": 14363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2617179066816218,
+      "learning_rate": 1.3534452934501795e-05,
+      "loss": 1.0801,
+      "step": 14364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.969385882153925,
+      "learning_rate": 1.353074701101214e-05,
+      "loss": 0.9688,
+      "step": 14365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546380960489055,
+      "learning_rate": 1.3527041406707139e-05,
+      "loss": 1.1211,
+      "step": 14366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5619847783361738,
+      "learning_rate": 1.3523336121689903e-05,
+      "loss": 0.425,
+      "step": 14367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8635601872569088,
+      "learning_rate": 1.3519631156063539e-05,
+      "loss": 0.7305,
+      "step": 14368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.004393548852534,
+      "learning_rate": 1.3515926509931178e-05,
+      "loss": 0.7178,
+      "step": 14369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.520900620332256,
+      "learning_rate": 1.3512222183395899e-05,
+      "loss": 0.9878,
+      "step": 14370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8999039123732602,
+      "learning_rate": 1.3508518176560808e-05,
+      "loss": 1.127,
+      "step": 14371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.209790798486905,
+      "learning_rate": 1.3504814489528978e-05,
+      "loss": 0.6094,
+      "step": 14372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.735525073066069,
+      "learning_rate": 1.3501111122403471e-05,
+      "loss": 1.4512,
+      "step": 14373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.851330485135923,
+      "learning_rate": 1.3497408075287369e-05,
+      "loss": 1.1348,
+      "step": 14374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.894100857965083,
+      "learning_rate": 1.349370534828372e-05,
+      "loss": 1.2168,
+      "step": 14375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0596225201938574,
+      "learning_rate": 1.3490002941495564e-05,
+      "loss": 0.9258,
+      "step": 14376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6477217255592658,
+      "learning_rate": 1.348630085502593e-05,
+      "loss": 0.5366,
+      "step": 14377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5895658599257017,
+      "learning_rate": 1.348259908897786e-05,
+      "loss": 1.3418,
+      "step": 14378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0637324441868268,
+      "learning_rate": 1.3478897643454375e-05,
+      "loss": 0.9697,
+      "step": 14379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3021807987990184,
+      "learning_rate": 1.347519651855848e-05,
+      "loss": 1.4102,
+      "step": 14380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8310498029989772,
+      "learning_rate": 1.3471495714393172e-05,
+      "loss": 1.0381,
+      "step": 14381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.262486683217834,
+      "learning_rate": 1.3467795231061441e-05,
+      "loss": 1.4102,
+      "step": 14382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5068022158814482,
+      "learning_rate": 1.3464095068666278e-05,
+      "loss": 1.3555,
+      "step": 14383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5171648134598015,
+      "learning_rate": 1.3460395227310657e-05,
+      "loss": 0.9521,
+      "step": 14384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3420741892986037,
+      "learning_rate": 1.3456695707097535e-05,
+      "loss": 0.9727,
+      "step": 14385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.307388824878052,
+      "learning_rate": 1.3452996508129866e-05,
+      "loss": 1.4277,
+      "step": 14386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.158240471632124,
+      "learning_rate": 1.3449297630510601e-05,
+      "loss": 0.8604,
+      "step": 14387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1791307839040215,
+      "learning_rate": 1.344559907434269e-05,
+      "loss": 1.0029,
+      "step": 14388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.562359925981687,
+      "learning_rate": 1.3441900839729054e-05,
+      "loss": 0.7773,
+      "step": 14389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3587425344214585,
+      "learning_rate": 1.3438202926772614e-05,
+      "loss": 0.9805,
+      "step": 14390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6494194618448486,
+      "learning_rate": 1.343450533557627e-05,
+      "loss": 1.5645,
+      "step": 14391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.565459931896019,
+      "learning_rate": 1.3430808066242939e-05,
+      "loss": 1.459,
+      "step": 14392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.465784456818029,
+      "learning_rate": 1.3427111118875513e-05,
+      "loss": 1.2539,
+      "step": 14393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5173857787812746,
+      "learning_rate": 1.3423414493576863e-05,
+      "loss": 1.3477,
+      "step": 14394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.258635163166679,
+      "learning_rate": 1.3419718190449877e-05,
+      "loss": 0.8926,
+      "step": 14395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.589044957331896,
+      "learning_rate": 1.3416022209597429e-05,
+      "loss": 0.8389,
+      "step": 14396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4304222067288537,
+      "learning_rate": 1.3412326551122365e-05,
+      "loss": 1.208,
+      "step": 14397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4203133747451413,
+      "learning_rate": 1.3408631215127538e-05,
+      "loss": 0.7842,
+      "step": 14398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5250000264384957,
+      "learning_rate": 1.3404936201715773e-05,
+      "loss": 1.3008,
+      "step": 14399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.078512042041175,
+      "learning_rate": 1.3401241510989924e-05,
+      "loss": 0.6685,
+      "step": 14400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7765331775531634,
+      "learning_rate": 1.3397547143052803e-05,
+      "loss": 0.792,
+      "step": 14401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.083436224891492,
+      "learning_rate": 1.339385309800722e-05,
+      "loss": 1.04,
+      "step": 14402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1540705643950178,
+      "learning_rate": 1.3390159375955969e-05,
+      "loss": 0.6865,
+      "step": 14403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2083129402034753,
+      "learning_rate": 1.338646597700187e-05,
+      "loss": 1.3916,
+      "step": 14404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.066742741290793,
+      "learning_rate": 1.338277290124768e-05,
+      "loss": 1.416,
+      "step": 14405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0643577390647962,
+      "learning_rate": 1.3379080148796208e-05,
+      "loss": 1.2285,
+      "step": 14406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4180321507931377,
+      "learning_rate": 1.33753877197502e-05,
+      "loss": 0.6973,
+      "step": 14407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.664013635629118,
+      "learning_rate": 1.337169561421241e-05,
+      "loss": 1.0063,
+      "step": 14408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.402005131685669,
+      "learning_rate": 1.3368003832285605e-05,
+      "loss": 1.5762,
+      "step": 14409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.455747041337193,
+      "learning_rate": 1.336431237407252e-05,
+      "loss": 1.2217,
+      "step": 14410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6078349611333302,
+      "learning_rate": 1.3360621239675885e-05,
+      "loss": 0.7983,
+      "step": 14411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3615233890393674,
+      "learning_rate": 1.335693042919841e-05,
+      "loss": 0.709,
+      "step": 14412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.848758937554115,
+      "learning_rate": 1.3353239942742835e-05,
+      "loss": 1.1562,
+      "step": 14413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.240377640730922,
+      "learning_rate": 1.3349549780411841e-05,
+      "loss": 1.3184,
+      "step": 14414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.677363307307325,
+      "learning_rate": 1.3345859942308141e-05,
+      "loss": 0.9834,
+      "step": 14415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3570912405035043,
+      "learning_rate": 1.3342170428534406e-05,
+      "loss": 0.4956,
+      "step": 14416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8221143213389315,
+      "learning_rate": 1.3338481239193335e-05,
+      "loss": 0.8613,
+      "step": 14417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.804940979960543,
+      "learning_rate": 1.3334792374387578e-05,
+      "loss": 0.9946,
+      "step": 14418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.821810744544125,
+      "learning_rate": 1.3331103834219805e-05,
+      "loss": 0.8584,
+      "step": 14419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0442335367433477,
+      "learning_rate": 1.3327415618792649e-05,
+      "loss": 1.4688,
+      "step": 14420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7781750257212483,
+      "learning_rate": 1.3323727728208773e-05,
+      "loss": 1.168,
+      "step": 14421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.90702877573194,
+      "learning_rate": 1.3320040162570804e-05,
+      "loss": 1.3906,
+      "step": 14422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8100867196273804,
+      "learning_rate": 1.331635292198135e-05,
+      "loss": 1.041,
+      "step": 14423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0913099434451525,
+      "learning_rate": 1.331266600654305e-05,
+      "loss": 1.1729,
+      "step": 14424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8757775919543738,
+      "learning_rate": 1.3308979416358486e-05,
+      "loss": 0.7636,
+      "step": 14425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5384317876499574,
+      "learning_rate": 1.3305293151530276e-05,
+      "loss": 0.6963,
+      "step": 14426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1638727759089313,
+      "learning_rate": 1.3301607212160997e-05,
+      "loss": 0.9104,
+      "step": 14427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0840043205597096,
+      "learning_rate": 1.3297921598353225e-05,
+      "loss": 1.0811,
+      "step": 14428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.285005727687948,
+      "learning_rate": 1.3294236310209524e-05,
+      "loss": 1.1377,
+      "step": 14429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.401984676583683,
+      "learning_rate": 1.3290551347832469e-05,
+      "loss": 0.8672,
+      "step": 14430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1438707679118245,
+      "learning_rate": 1.3286866711324603e-05,
+      "loss": 0.8223,
+      "step": 14431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.220025373359372,
+      "learning_rate": 1.3283182400788458e-05,
+      "loss": 1.4414,
+      "step": 14432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.14184273412773,
+      "learning_rate": 1.3279498416326577e-05,
+      "loss": 0.7385,
+      "step": 14433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.882240005500905,
+      "learning_rate": 1.3275814758041493e-05,
+      "loss": 1.4844,
+      "step": 14434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7454317878375183,
+      "learning_rate": 1.3272131426035713e-05,
+      "loss": 1.0898,
+      "step": 14435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.936358453234439,
+      "learning_rate": 1.3268448420411738e-05,
+      "loss": 0.3633,
+      "step": 14436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2770783415306592,
+      "learning_rate": 1.326476574127206e-05,
+      "loss": 1.2578,
+      "step": 14437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6478857855467974,
+      "learning_rate": 1.3261083388719187e-05,
+      "loss": 1.2734,
+      "step": 14438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.625120886789324,
+      "learning_rate": 1.3257401362855581e-05,
+      "loss": 0.7544,
+      "step": 14439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0163594172114854,
+      "learning_rate": 1.3253719663783715e-05,
+      "loss": 0.9337,
+      "step": 14440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.129370714696084,
+      "learning_rate": 1.3250038291606038e-05,
+      "loss": 1.043,
+      "step": 14441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.032593419585082,
+      "learning_rate": 1.3246357246425012e-05,
+      "loss": 1.0518,
+      "step": 14442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4461368231695495,
+      "learning_rate": 1.324267652834309e-05,
+      "loss": 1.2373,
+      "step": 14443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.826400483961844,
+      "learning_rate": 1.3238996137462694e-05,
+      "loss": 1.0957,
+      "step": 14444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.919269072877805,
+      "learning_rate": 1.3235316073886244e-05,
+      "loss": 1.3896,
+      "step": 14445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.933116970608067,
+      "learning_rate": 1.3231636337716152e-05,
+      "loss": 1.1328,
+      "step": 14446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.708368848910138,
+      "learning_rate": 1.3227956929054841e-05,
+      "loss": 0.9902,
+      "step": 14447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6776710686410232,
+      "learning_rate": 1.3224277848004692e-05,
+      "loss": 1.4531,
+      "step": 14448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.380864026526717,
+      "learning_rate": 1.3220599094668101e-05,
+      "loss": 1.4102,
+      "step": 14449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5418467905951716,
+      "learning_rate": 1.3216920669147431e-05,
+      "loss": 0.7019,
+      "step": 14450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.750273604220367,
+      "learning_rate": 1.3213242571545058e-05,
+      "loss": 0.9385,
+      "step": 14451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5052827748768207,
+      "learning_rate": 1.3209564801963359e-05,
+      "loss": 0.9805,
+      "step": 14452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.392721340627637,
+      "learning_rate": 1.320588736050467e-05,
+      "loss": 0.7236,
+      "step": 14453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7967944980454478,
+      "learning_rate": 1.3202210247271326e-05,
+      "loss": 1.334,
+      "step": 14454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.130733875369906,
+      "learning_rate": 1.3198533462365678e-05,
+      "loss": 0.8047,
+      "step": 14455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.946909762541709,
+      "learning_rate": 1.3194857005890037e-05,
+      "loss": 1.1719,
+      "step": 14456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.288579063011595,
+      "learning_rate": 1.3191180877946724e-05,
+      "loss": 1.0469,
+      "step": 14457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.18793997659453,
+      "learning_rate": 1.3187505078638026e-05,
+      "loss": 1.0425,
+      "step": 14458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.847489968299778,
+      "learning_rate": 1.3183829608066264e-05,
+      "loss": 1.1797,
+      "step": 14459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6337541203767603,
+      "learning_rate": 1.3180154466333705e-05,
+      "loss": 0.8765,
+      "step": 14460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.803345721540351,
+      "learning_rate": 1.3176479653542645e-05,
+      "loss": 1.4727,
+      "step": 14461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 26.759106948616576,
+      "learning_rate": 1.3172805169795338e-05,
+      "loss": 0.8916,
+      "step": 14462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.732466793371714,
+      "learning_rate": 1.3169131015194045e-05,
+      "loss": 1.5352,
+      "step": 14463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.225273212791406,
+      "learning_rate": 1.3165457189841024e-05,
+      "loss": 1.0352,
+      "step": 14464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.078289857219473,
+      "learning_rate": 1.3161783693838514e-05,
+      "loss": 0.9297,
+      "step": 14465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.849614684539482,
+      "learning_rate": 1.3158110527288748e-05,
+      "loss": 0.8701,
+      "step": 14466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.946483520042175,
+      "learning_rate": 1.315443769029393e-05,
+      "loss": 0.9922,
+      "step": 14467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6450563439688481,
+      "learning_rate": 1.3150765182956302e-05,
+      "loss": 0.6567,
+      "step": 14468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.857860878329739,
+      "learning_rate": 1.3147093005378053e-05,
+      "loss": 0.6958,
+      "step": 14469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6863996512743995,
+      "learning_rate": 1.314342115766137e-05,
+      "loss": 1.2041,
+      "step": 14470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.590385083414343,
+      "learning_rate": 1.313974963990845e-05,
+      "loss": 1.8906,
+      "step": 14471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.200761435378459,
+      "learning_rate": 1.313607845222148e-05,
+      "loss": 1.1396,
+      "step": 14472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.836005153034072,
+      "learning_rate": 1.3132407594702617e-05,
+      "loss": 1.3496,
+      "step": 14473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3040806892822543,
+      "learning_rate": 1.3128737067454017e-05,
+      "loss": 0.8525,
+      "step": 14474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.5890413210848555,
+      "learning_rate": 1.3125066870577823e-05,
+      "loss": 1.6089,
+      "step": 14475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1072943545585736,
+      "learning_rate": 1.3121397004176194e-05,
+      "loss": 0.7803,
+      "step": 14476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.196346240009044,
+      "learning_rate": 1.3117727468351254e-05,
+      "loss": 1.3828,
+      "step": 14477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7115717021645311,
+      "learning_rate": 1.3114058263205118e-05,
+      "loss": 1.0107,
+      "step": 14478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4063747423213573,
+      "learning_rate": 1.3110389388839889e-05,
+      "loss": 0.9277,
+      "step": 14479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4539263731813135,
+      "learning_rate": 1.310672084535769e-05,
+      "loss": 0.8057,
+      "step": 14480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.998887264928205,
+      "learning_rate": 1.3103052632860613e-05,
+      "loss": 0.9438,
+      "step": 14481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1779416109705787,
+      "learning_rate": 1.3099384751450739e-05,
+      "loss": 1.2559,
+      "step": 14482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4922717466155855,
+      "learning_rate": 1.309571720123014e-05,
+      "loss": 0.9624,
+      "step": 14483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.171545002124436,
+      "learning_rate": 1.309204998230088e-05,
+      "loss": 0.4082,
+      "step": 14484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.167056509917863,
+      "learning_rate": 1.3088383094765028e-05,
+      "loss": 0.9374,
+      "step": 14485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.50692241565261,
+      "learning_rate": 1.3084716538724626e-05,
+      "loss": 0.9561,
+      "step": 14486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.130867224431077,
+      "learning_rate": 1.3081050314281711e-05,
+      "loss": 1.1113,
+      "step": 14487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4184208017520286,
+      "learning_rate": 1.3077384421538305e-05,
+      "loss": 1.541,
+      "step": 14488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.5578065423341,
+      "learning_rate": 1.3073718860596437e-05,
+      "loss": 1.8711,
+      "step": 14489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.623863246558289,
+      "learning_rate": 1.3070053631558127e-05,
+      "loss": 1.5996,
+      "step": 14490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9479523421825844,
+      "learning_rate": 1.306638873452537e-05,
+      "loss": 0.812,
+      "step": 14491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.203295298445488,
+      "learning_rate": 1.3062724169600143e-05,
+      "loss": 1.0747,
+      "step": 14492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.003750622154896,
+      "learning_rate": 1.3059059936884455e-05,
+      "loss": 0.9688,
+      "step": 14493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6746322344018956,
+      "learning_rate": 1.3055396036480267e-05,
+      "loss": 0.9326,
+      "step": 14494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2434300259339466,
+      "learning_rate": 1.3051732468489544e-05,
+      "loss": 1.1611,
+      "step": 14495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.338953375098859,
+      "learning_rate": 1.3048069233014232e-05,
+      "loss": 1.4375,
+      "step": 14496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4204975764201984,
+      "learning_rate": 1.3044406330156288e-05,
+      "loss": 1.5254,
+      "step": 14497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.574935270591546,
+      "learning_rate": 1.3040743760017659e-05,
+      "loss": 1.5859,
+      "step": 14498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.162329877127891,
+      "learning_rate": 1.3037081522700265e-05,
+      "loss": 1.7246,
+      "step": 14499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.247337885728765,
+      "learning_rate": 1.303341961830602e-05,
+      "loss": 0.7297,
+      "step": 14500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.624544482094438,
+      "learning_rate": 1.3029758046936823e-05,
+      "loss": 1.1953,
+      "step": 14501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.366361266118229,
+      "learning_rate": 1.3026096808694601e-05,
+      "loss": 1.3232,
+      "step": 14502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.216613060170188,
+      "learning_rate": 1.302243590368123e-05,
+      "loss": 1.1621,
+      "step": 14503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8184275589340633,
+      "learning_rate": 1.301877533199859e-05,
+      "loss": 1.0596,
+      "step": 14504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.335979289635328,
+      "learning_rate": 1.3015115093748542e-05,
+      "loss": 0.8691,
+      "step": 14505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.621077285919556,
+      "learning_rate": 1.3011455189032976e-05,
+      "loss": 0.834,
+      "step": 14506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4150228655428028,
+      "learning_rate": 1.300779561795372e-05,
+      "loss": 1.0547,
+      "step": 14507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.350292483856865,
+      "learning_rate": 1.3004136380612642e-05,
+      "loss": 1.0742,
+      "step": 14508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9283721394482116,
+      "learning_rate": 1.300047747711155e-05,
+      "loss": 0.7446,
+      "step": 14509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3905123017119636,
+      "learning_rate": 1.29968189075523e-05,
+      "loss": 0.877,
+      "step": 14510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1292878298068625,
+      "learning_rate": 1.2993160672036691e-05,
+      "loss": 1.2578,
+      "step": 14511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.232515399421187,
+      "learning_rate": 1.298950277066653e-05,
+      "loss": 1.4961,
+      "step": 14512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.041568492077123,
+      "learning_rate": 1.2985845203543611e-05,
+      "loss": 1.1348,
+      "step": 14513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.377776318413361,
+      "learning_rate": 1.2982187970769738e-05,
+      "loss": 1.3096,
+      "step": 14514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5354803080996766,
+      "learning_rate": 1.2978531072446682e-05,
+      "loss": 0.9578,
+      "step": 14515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3940462053663367,
+      "learning_rate": 1.2974874508676201e-05,
+      "loss": 1.248,
+      "step": 14516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8245300712314987,
+      "learning_rate": 1.297121827956008e-05,
+      "loss": 0.7148,
+      "step": 14517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2753558373145433,
+      "learning_rate": 1.2967562385200044e-05,
+      "loss": 1.3379,
+      "step": 14518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.947088911065084,
+      "learning_rate": 1.296390682569786e-05,
+      "loss": 1.1992,
+      "step": 14519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.019897596631486,
+      "learning_rate": 1.2960251601155249e-05,
+      "loss": 1.0977,
+      "step": 14520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.716243340450623,
+      "learning_rate": 1.2956596711673934e-05,
+      "loss": 1.1094,
+      "step": 14521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.48222140827939,
+      "learning_rate": 1.2952942157355619e-05,
+      "loss": 1.3848,
+      "step": 14522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.6310795795428845,
+      "learning_rate": 1.2949287938302029e-05,
+      "loss": 1.0361,
+      "step": 14523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4599730524277406,
+      "learning_rate": 1.2945634054614853e-05,
+      "loss": 1.2002,
+      "step": 14524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0797581092175896,
+      "learning_rate": 1.2941980506395762e-05,
+      "loss": 0.8564,
+      "step": 14525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.184844584251121,
+      "learning_rate": 1.2938327293746455e-05,
+      "loss": 1.5801,
+      "step": 14526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.887843427836123,
+      "learning_rate": 1.2934674416768578e-05,
+      "loss": 1.1641,
+      "step": 14527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.684239380990858,
+      "learning_rate": 1.2931021875563812e-05,
+      "loss": 0.8672,
+      "step": 14528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7172058489852926,
+      "learning_rate": 1.2927369670233794e-05,
+      "loss": 1.2363,
+      "step": 14529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9236891333238046,
+      "learning_rate": 1.2923717800880153e-05,
+      "loss": 0.7158,
+      "step": 14530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.476128506092538,
+      "learning_rate": 1.2920066267604536e-05,
+      "loss": 1.4844,
+      "step": 14531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3997445288427968,
+      "learning_rate": 1.291641507050856e-05,
+      "loss": 0.6973,
+      "step": 14532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.20503703806385,
+      "learning_rate": 1.2912764209693834e-05,
+      "loss": 1.1582,
+      "step": 14533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8969341337988495,
+      "learning_rate": 1.2909113685261947e-05,
+      "loss": 0.9117,
+      "step": 14534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5988190536535654,
+      "learning_rate": 1.2905463497314508e-05,
+      "loss": 1.0205,
+      "step": 14535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.2376748915444775,
+      "learning_rate": 1.2901813645953104e-05,
+      "loss": 2.1958,
+      "step": 14536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.148998617243688,
+      "learning_rate": 1.28981641312793e-05,
+      "loss": 0.9688,
+      "step": 14537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9506024970431437,
+      "learning_rate": 1.289451495339466e-05,
+      "loss": 1.1831,
+      "step": 14538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.579779524514697,
+      "learning_rate": 1.2890866112400734e-05,
+      "loss": 1.1436,
+      "step": 14539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.601738451257159,
+      "learning_rate": 1.2887217608399083e-05,
+      "loss": 0.9805,
+      "step": 14540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.479755771234511,
+      "learning_rate": 1.2883569441491233e-05,
+      "loss": 0.8408,
+      "step": 14541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.818884459780222,
+      "learning_rate": 1.2879921611778717e-05,
+      "loss": 0.5894,
+      "step": 14542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.643798005229608,
+      "learning_rate": 1.2876274119363035e-05,
+      "loss": 1.1465,
+      "step": 14543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.524053887629626,
+      "learning_rate": 1.2872626964345713e-05,
+      "loss": 1.0381,
+      "step": 14544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.36930495633313,
+      "learning_rate": 1.2868980146828252e-05,
+      "loss": 0.8281,
+      "step": 14545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4879186536731708,
+      "learning_rate": 1.2865333666912132e-05,
+      "loss": 1.0498,
+      "step": 14546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.844576015005993,
+      "learning_rate": 1.2861687524698842e-05,
+      "loss": 0.9922,
+      "step": 14547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3176323339519844,
+      "learning_rate": 1.2858041720289835e-05,
+      "loss": 1.3398,
+      "step": 14548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5500320806542263,
+      "learning_rate": 1.2854396253786593e-05,
+      "loss": 1.1406,
+      "step": 14549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5443272788182245,
+      "learning_rate": 1.2850751125290559e-05,
+      "loss": 1.0332,
+      "step": 14550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1586035083636053,
+      "learning_rate": 1.2847106334903164e-05,
+      "loss": 0.9219,
+      "step": 14551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8758504611085036,
+      "learning_rate": 1.2843461882725865e-05,
+      "loss": 0.9297,
+      "step": 14552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.510639629250169,
+      "learning_rate": 1.2839817768860063e-05,
+      "loss": 1.2715,
+      "step": 14553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3100212834356175,
+      "learning_rate": 1.2836173993407188e-05,
+      "loss": 1.1133,
+      "step": 14554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.467234273021527,
+      "learning_rate": 1.2832530556468642e-05,
+      "loss": 1.6973,
+      "step": 14555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.095595045049132,
+      "learning_rate": 1.2828887458145806e-05,
+      "loss": 1.4199,
+      "step": 14556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7212593487255055,
+      "learning_rate": 1.282524469854009e-05,
+      "loss": 1.0391,
+      "step": 14557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.629543822290929,
+      "learning_rate": 1.2821602277752854e-05,
+      "loss": 0.9141,
+      "step": 14558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.403734849553758,
+      "learning_rate": 1.281796019588547e-05,
+      "loss": 1.4199,
+      "step": 14559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.721645085361537,
+      "learning_rate": 1.2814318453039285e-05,
+      "loss": 1.0732,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6967205566752686,
+      "learning_rate": 1.2810677049315667e-05,
+      "loss": 0.8159,
+      "step": 14561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.054560723647281,
+      "learning_rate": 1.2807035984815933e-05,
+      "loss": 1.5664,
+      "step": 14562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.433606201604526,
+      "learning_rate": 1.2803395259641433e-05,
+      "loss": 0.6841,
+      "step": 14563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.78342483025389,
+      "learning_rate": 1.279975487389348e-05,
+      "loss": 1.1416,
+      "step": 14564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.920738452294895,
+      "learning_rate": 1.2796114827673372e-05,
+      "loss": 0.9053,
+      "step": 14565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.518087851902932,
+      "learning_rate": 1.2792475121082428e-05,
+      "loss": 1.5454,
+      "step": 14566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.563104544759697,
+      "learning_rate": 1.2788835754221934e-05,
+      "loss": 0.9092,
+      "step": 14567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.01890551559418,
+      "learning_rate": 1.2785196727193169e-05,
+      "loss": 0.877,
+      "step": 14568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.577433267907978,
+      "learning_rate": 1.2781558040097396e-05,
+      "loss": 0.9062,
+      "step": 14569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8336027428689667,
+      "learning_rate": 1.27779196930359e-05,
+      "loss": 0.7266,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.262822617736783,
+      "learning_rate": 1.2774281686109913e-05,
+      "loss": 1.7539,
+      "step": 14571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.88021356615609,
+      "learning_rate": 1.2770644019420697e-05,
+      "loss": 1.1094,
+      "step": 14572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8908693139979342,
+      "learning_rate": 1.2767006693069472e-05,
+      "loss": 0.6738,
+      "step": 14573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.471255129942444,
+      "learning_rate": 1.276336970715748e-05,
+      "loss": 0.8965,
+      "step": 14574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0823764765307318,
+      "learning_rate": 1.2759733061785928e-05,
+      "loss": 1.2744,
+      "step": 14575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.644890584719746,
+      "learning_rate": 1.275609675705602e-05,
+      "loss": 1.1299,
+      "step": 14576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.284052864458347,
+      "learning_rate": 1.2752460793068948e-05,
+      "loss": 0.8027,
+      "step": 14577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3978046986554777,
+      "learning_rate": 1.2748825169925915e-05,
+      "loss": 0.8281,
+      "step": 14578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.072735245516295,
+      "learning_rate": 1.2745189887728087e-05,
+      "loss": 0.7949,
+      "step": 14579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.421172187430408,
+      "learning_rate": 1.274155494657664e-05,
+      "loss": 1.4004,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.610894223487469,
+      "learning_rate": 1.2737920346572716e-05,
+      "loss": 1.2139,
+      "step": 14581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.926307240848395,
+      "learning_rate": 1.2734286087817477e-05,
+      "loss": 1.6777,
+      "step": 14582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0580274659364775,
+      "learning_rate": 1.2730652170412072e-05,
+      "loss": 1.2734,
+      "step": 14583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8777452716546932,
+      "learning_rate": 1.2727018594457624e-05,
+      "loss": 0.7031,
+      "step": 14584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.258273735051218,
+      "learning_rate": 1.272338536005525e-05,
+      "loss": 1.4277,
+      "step": 14585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5647220165101725,
+      "learning_rate": 1.2719752467306056e-05,
+      "loss": 1.3516,
+      "step": 14586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5659966959614406,
+      "learning_rate": 1.271611991631116e-05,
+      "loss": 1.8047,
+      "step": 14587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3511833823144737,
+      "learning_rate": 1.2712487707171645e-05,
+      "loss": 1.1729,
+      "step": 14588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9042137026455523,
+      "learning_rate": 1.2708855839988598e-05,
+      "loss": 1.752,
+      "step": 14589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.192205110072711,
+      "learning_rate": 1.2705224314863067e-05,
+      "loss": 0.5144,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.238598546411538,
+      "learning_rate": 1.2701593131896162e-05,
+      "loss": 1.0654,
+      "step": 14591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.819399022664054,
+      "learning_rate": 1.2697962291188908e-05,
+      "loss": 0.7686,
+      "step": 14592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2566751433550905,
+      "learning_rate": 1.2694331792842357e-05,
+      "loss": 1.1953,
+      "step": 14593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.508485144265593,
+      "learning_rate": 1.2690701636957533e-05,
+      "loss": 1.1924,
+      "step": 14594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.409160329241835,
+      "learning_rate": 1.2687071823635483e-05,
+      "loss": 1.1758,
+      "step": 14595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.3068387167906375,
+      "learning_rate": 1.2683442352977209e-05,
+      "loss": 0.8359,
+      "step": 14596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.781540244741693,
+      "learning_rate": 1.267981322508372e-05,
+      "loss": 1.1426,
+      "step": 14597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6285009199216676,
+      "learning_rate": 1.2676184440056005e-05,
+      "loss": 1.1621,
+      "step": 14598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3497493080738026,
+      "learning_rate": 1.2672555997995059e-05,
+      "loss": 0.8198,
+      "step": 14599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2026127103198485,
+      "learning_rate": 1.266892789900187e-05,
+      "loss": 1.3867,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.298960977635478,
+      "learning_rate": 1.26653001431774e-05,
+      "loss": 1.4609,
+      "step": 14601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.0420255731059616,
+      "learning_rate": 1.2661672730622603e-05,
+      "loss": 1.6152,
+      "step": 14602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0121355236133978,
+      "learning_rate": 1.2658045661438424e-05,
+      "loss": 0.9697,
+      "step": 14603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3230673109624282,
+      "learning_rate": 1.2654418935725821e-05,
+      "loss": 1.4238,
+      "step": 14604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.892803355889719,
+      "learning_rate": 1.265079255358571e-05,
+      "loss": 0.8752,
+      "step": 14605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8802338105236687,
+      "learning_rate": 1.2647166515119013e-05,
+      "loss": 1.2207,
+      "step": 14606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0645162957089496,
+      "learning_rate": 1.2643540820426636e-05,
+      "loss": 0.7432,
+      "step": 14607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.888431037714131,
+      "learning_rate": 1.2639915469609498e-05,
+      "loss": 0.9385,
+      "step": 14608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6925224569618535,
+      "learning_rate": 1.263629046276847e-05,
+      "loss": 1.0273,
+      "step": 14609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.797406320326772,
+      "learning_rate": 1.2632665800004456e-05,
+      "loss": 0.4351,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.369954421494417,
+      "learning_rate": 1.2629041481418307e-05,
+      "loss": 1.2744,
+      "step": 14611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.65682131224376,
+      "learning_rate": 1.262541750711091e-05,
+      "loss": 0.8413,
+      "step": 14612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4707769450743204,
+      "learning_rate": 1.2621793877183103e-05,
+      "loss": 0.8914,
+      "step": 14613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4615562006503997,
+      "learning_rate": 1.2618170591735731e-05,
+      "loss": 1.0254,
+      "step": 14614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.128685840081086,
+      "learning_rate": 1.2614547650869623e-05,
+      "loss": 0.7971,
+      "step": 14615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.0217637483844735,
+      "learning_rate": 1.2610925054685624e-05,
+      "loss": 1.2734,
+      "step": 14616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.6418256137738005,
+      "learning_rate": 1.2607302803284535e-05,
+      "loss": 1.2578,
+      "step": 14617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3966383837141345,
+      "learning_rate": 1.2603680896767155e-05,
+      "loss": 1.3711,
+      "step": 14618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9087769382280224,
+      "learning_rate": 1.2600059335234298e-05,
+      "loss": 0.646,
+      "step": 14619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.655591987021217,
+      "learning_rate": 1.2596438118786732e-05,
+      "loss": 0.915,
+      "step": 14620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7224794808175132,
+      "learning_rate": 1.2592817247525254e-05,
+      "loss": 0.8574,
+      "step": 14621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8242188344193018,
+      "learning_rate": 1.2589196721550622e-05,
+      "loss": 1.3613,
+      "step": 14622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.986020258820088,
+      "learning_rate": 1.258557654096359e-05,
+      "loss": 1.4844,
+      "step": 14623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.020154644372989,
+      "learning_rate": 1.25819567058649e-05,
+      "loss": 1.2461,
+      "step": 14624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.947768156621597,
+      "learning_rate": 1.2578337216355312e-05,
+      "loss": 1.1006,
+      "step": 14625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.25455257992863,
+      "learning_rate": 1.257471807253554e-05,
+      "loss": 0.9727,
+      "step": 14626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.034465853649992,
+      "learning_rate": 1.2571099274506295e-05,
+      "loss": 0.8838,
+      "step": 14627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.911601458581453,
+      "learning_rate": 1.25674808223683e-05,
+      "loss": 0.9043,
+      "step": 14628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.455875872172385,
+      "learning_rate": 1.256386271622226e-05,
+      "loss": 0.9912,
+      "step": 14629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2238799759931163,
+      "learning_rate": 1.2560244956168859e-05,
+      "loss": 0.7109,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.506115447411494,
+      "learning_rate": 1.2556627542308777e-05,
+      "loss": 1.3945,
+      "step": 14631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.101530393901187,
+      "learning_rate": 1.2553010474742674e-05,
+      "loss": 1.2412,
+      "step": 14632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6984406383432096,
+      "learning_rate": 1.2549393753571232e-05,
+      "loss": 0.6829,
+      "step": 14633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.246141940719439,
+      "learning_rate": 1.2545777378895096e-05,
+      "loss": 0.8594,
+      "step": 14634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.240676757453762,
+      "learning_rate": 1.2542161350814902e-05,
+      "loss": 1.4639,
+      "step": 14635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.211139511330243,
+      "learning_rate": 1.2538545669431279e-05,
+      "loss": 1.1523,
+      "step": 14636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2324294576574526,
+      "learning_rate": 1.2534930334844853e-05,
+      "loss": 1.1191,
+      "step": 14637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.16272647531706,
+      "learning_rate": 1.2531315347156253e-05,
+      "loss": 0.4937,
+      "step": 14638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7572769639262464,
+      "learning_rate": 1.2527700706466072e-05,
+      "loss": 0.6226,
+      "step": 14639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0220796602368933,
+      "learning_rate": 1.2524086412874903e-05,
+      "loss": 0.8242,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.263252464841778,
+      "learning_rate": 1.252047246648332e-05,
+      "loss": 0.6387,
+      "step": 14641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.535496199610831,
+      "learning_rate": 1.2516858867391917e-05,
+      "loss": 1.8359,
+      "step": 14642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.271373633966861,
+      "learning_rate": 1.251324561570125e-05,
+      "loss": 0.8232,
+      "step": 14643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.427067463267645,
+      "learning_rate": 1.2509632711511874e-05,
+      "loss": 1.0502,
+      "step": 14644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3155416126140587,
+      "learning_rate": 1.2506020154924325e-05,
+      "loss": 1.209,
+      "step": 14645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.725378329488218,
+      "learning_rate": 1.2502407946039149e-05,
+      "loss": 0.7642,
+      "step": 14646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.243243737564676,
+      "learning_rate": 1.2498796084956882e-05,
+      "loss": 1.2041,
+      "step": 14647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.634615860649293,
+      "learning_rate": 1.2495184571778029e-05,
+      "loss": 1.4199,
+      "step": 14648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8715389733421004,
+      "learning_rate": 1.2491573406603091e-05,
+      "loss": 0.8708,
+      "step": 14649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.59222624728519,
+      "learning_rate": 1.2487962589532584e-05,
+      "loss": 1.4043,
+      "step": 14650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.382898958394915,
+      "learning_rate": 1.2484352120666981e-05,
+      "loss": 0.9736,
+      "step": 14651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.700624184145732,
+      "learning_rate": 1.2480742000106766e-05,
+      "loss": 1.2539,
+      "step": 14652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1754254199119023,
+      "learning_rate": 1.2477132227952392e-05,
+      "loss": 0.8311,
+      "step": 14653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.257150559906229,
+      "learning_rate": 1.2473522804304343e-05,
+      "loss": 0.9883,
+      "step": 14654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5480744306723793,
+      "learning_rate": 1.2469913729263044e-05,
+      "loss": 0.7665,
+      "step": 14655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7937517305691442,
+      "learning_rate": 1.2466305002928955e-05,
+      "loss": 0.8147,
+      "step": 14656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6917593973768277,
+      "learning_rate": 1.2462696625402493e-05,
+      "loss": 1.4902,
+      "step": 14657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5892534000278276,
+      "learning_rate": 1.2459088596784071e-05,
+      "loss": 0.6045,
+      "step": 14658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6695872545717116,
+      "learning_rate": 1.245548091717412e-05,
+      "loss": 0.8379,
+      "step": 14659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2182051298533936,
+      "learning_rate": 1.2451873586673027e-05,
+      "loss": 0.7529,
+      "step": 14660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2017257233135865,
+      "learning_rate": 1.244826660538118e-05,
+      "loss": 1.3457,
+      "step": 14661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.784257612740165,
+      "learning_rate": 1.244465997339896e-05,
+      "loss": 0.7734,
+      "step": 14662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.346000270708966,
+      "learning_rate": 1.2441053690826748e-05,
+      "loss": 0.8916,
+      "step": 14663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.35303373509521,
+      "learning_rate": 1.2437447757764891e-05,
+      "loss": 1.1113,
+      "step": 14664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8935743924040802,
+      "learning_rate": 1.2433842174313756e-05,
+      "loss": 0.788,
+      "step": 14665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.738795429446659,
+      "learning_rate": 1.2430236940573669e-05,
+      "loss": 0.8066,
+      "step": 14666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.660208680564668,
+      "learning_rate": 1.2426632056644983e-05,
+      "loss": 0.8838,
+      "step": 14667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.71867649498585,
+      "learning_rate": 1.2423027522628005e-05,
+      "loss": 1.3682,
+      "step": 14668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.273998492788119,
+      "learning_rate": 1.2419423338623052e-05,
+      "loss": 1.0273,
+      "step": 14669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.562762848098354,
+      "learning_rate": 1.2415819504730417e-05,
+      "loss": 1.2275,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.575724900384875,
+      "learning_rate": 1.241221602105041e-05,
+      "loss": 0.7356,
+      "step": 14671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7449731831424447,
+      "learning_rate": 1.2408612887683308e-05,
+      "loss": 0.9453,
+      "step": 14672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.776268524996521,
+      "learning_rate": 1.2405010104729372e-05,
+      "loss": 1.521,
+      "step": 14673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4266876640389525,
+      "learning_rate": 1.240140767228889e-05,
+      "loss": 1.2949,
+      "step": 14674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.964809215273308,
+      "learning_rate": 1.2397805590462092e-05,
+      "loss": 1.8369,
+      "step": 14675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.945521585192968,
+      "learning_rate": 1.2394203859349244e-05,
+      "loss": 0.4482,
+      "step": 14676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.297616552247204,
+      "learning_rate": 1.239060247905057e-05,
+      "loss": 1.0986,
+      "step": 14677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.120377746113823,
+      "learning_rate": 1.2387001449666296e-05,
+      "loss": 1.4736,
+      "step": 14678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.6952552220663994,
+      "learning_rate": 1.238340077129663e-05,
+      "loss": 1.2832,
+      "step": 14679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7324383554139686,
+      "learning_rate": 1.2379800444041793e-05,
+      "loss": 0.9844,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2159266171062395,
+      "learning_rate": 1.237620046800197e-05,
+      "loss": 1.0547,
+      "step": 14681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.732733789826237,
+      "learning_rate": 1.237260084327735e-05,
+      "loss": 1.6543,
+      "step": 14682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3400936852961967,
+      "learning_rate": 1.2369001569968097e-05,
+      "loss": 0.9858,
+      "step": 14683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.515156416293555,
+      "learning_rate": 1.2365402648174388e-05,
+      "loss": 1.1953,
+      "step": 14684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1143385026490025,
+      "learning_rate": 1.2361804077996391e-05,
+      "loss": 1.2852,
+      "step": 14685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4433981492428356,
+      "learning_rate": 1.2358205859534242e-05,
+      "loss": 0.9087,
+      "step": 14686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.743732810259602,
+      "learning_rate": 1.2354607992888065e-05,
+      "loss": 1.1504,
+      "step": 14687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.249930724713731,
+      "learning_rate": 1.2351010478158012e-05,
+      "loss": 0.9138,
+      "step": 14688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9904922184819793,
+      "learning_rate": 1.2347413315444187e-05,
+      "loss": 1.0645,
+      "step": 14689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.418418380951763,
+      "learning_rate": 1.2343816504846698e-05,
+      "loss": 1.0142,
+      "step": 14690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.477076867314315,
+      "learning_rate": 1.2340220046465634e-05,
+      "loss": 1.6992,
+      "step": 14691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.645643814851476,
+      "learning_rate": 1.2336623940401094e-05,
+      "loss": 0.8662,
+      "step": 14692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.65001283408221,
+      "learning_rate": 1.2333028186753162e-05,
+      "loss": 1.1777,
+      "step": 14693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5027043973848966,
+      "learning_rate": 1.23294327856219e-05,
+      "loss": 0.8301,
+      "step": 14694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1718499710507007,
+      "learning_rate": 1.2325837737107365e-05,
+      "loss": 0.8926,
+      "step": 14695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0952579406828686,
+      "learning_rate": 1.2322243041309595e-05,
+      "loss": 1.6055,
+      "step": 14696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9405329561135563,
+      "learning_rate": 1.2318648698328653e-05,
+      "loss": 1.0113,
+      "step": 14697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8764682835157034,
+      "learning_rate": 1.2315054708264553e-05,
+      "loss": 0.8633,
+      "step": 14698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7703763110250177,
+      "learning_rate": 1.2311461071217314e-05,
+      "loss": 0.7881,
+      "step": 14699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.594683974230358,
+      "learning_rate": 1.2307867787286942e-05,
+      "loss": 1.373,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.43234882382878,
+      "learning_rate": 1.2304274856573443e-05,
+      "loss": 1.0645,
+      "step": 14701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.852131157179899,
+      "learning_rate": 1.2300682279176814e-05,
+      "loss": 1.418,
+      "step": 14702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.348558014100641,
+      "learning_rate": 1.2297090055197028e-05,
+      "loss": 1.4668,
+      "step": 14703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6925685017407006,
+      "learning_rate": 1.2293498184734053e-05,
+      "loss": 1.2783,
+      "step": 14704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7024994054576634,
+      "learning_rate": 1.2289906667887843e-05,
+      "loss": 0.8301,
+      "step": 14705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4876434131475844,
+      "learning_rate": 1.2286315504758364e-05,
+      "loss": 1.3145,
+      "step": 14706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.55553458501026,
+      "learning_rate": 1.228272469544555e-05,
+      "loss": 0.8076,
+      "step": 14707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.084147395985845,
+      "learning_rate": 1.227913424004932e-05,
+      "loss": 0.8428,
+      "step": 14708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.847080713682284,
+      "learning_rate": 1.2275544138669617e-05,
+      "loss": 0.7708,
+      "step": 14709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.056865881276652,
+      "learning_rate": 1.2271954391406328e-05,
+      "loss": 1.5488,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.094103243919505,
+      "learning_rate": 1.2268364998359377e-05,
+      "loss": 1.5742,
+      "step": 14711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.0687757611355035,
+      "learning_rate": 1.2264775959628647e-05,
+      "loss": 1.2422,
+      "step": 14712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.866657042916412,
+      "learning_rate": 1.2261187275314007e-05,
+      "loss": 1.4521,
+      "step": 14713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.961886674084272,
+      "learning_rate": 1.2257598945515348e-05,
+      "loss": 0.5356,
+      "step": 14714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.998081547539733,
+      "learning_rate": 1.2254010970332525e-05,
+      "loss": 0.9688,
+      "step": 14715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.590078950090772,
+      "learning_rate": 1.2250423349865387e-05,
+      "loss": 1.1836,
+      "step": 14716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1863984841326136,
+      "learning_rate": 1.2246836084213763e-05,
+      "loss": 1.3184,
+      "step": 14717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5450173986251556,
+      "learning_rate": 1.2243249173477513e-05,
+      "loss": 0.5,
+      "step": 14718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8613553440469484,
+      "learning_rate": 1.2239662617756445e-05,
+      "loss": 0.7461,
+      "step": 14719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.203860453095667,
+      "learning_rate": 1.2236076417150361e-05,
+      "loss": 1.0156,
+      "step": 14720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5227962659419574,
+      "learning_rate": 1.2232490571759084e-05,
+      "loss": 1.0752,
+      "step": 14721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.090343722137215,
+      "learning_rate": 1.2228905081682387e-05,
+      "loss": 0.8428,
+      "step": 14722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.641278872678802,
+      "learning_rate": 1.2225319947020072e-05,
+      "loss": 1.3008,
+      "step": 14723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.941892968402957,
+      "learning_rate": 1.2221735167871904e-05,
+      "loss": 2.2871,
+      "step": 14724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0443706771538985,
+      "learning_rate": 1.2218150744337642e-05,
+      "loss": 1.1738,
+      "step": 14725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.643027752999389,
+      "learning_rate": 1.2214566676517031e-05,
+      "loss": 0.8403,
+      "step": 14726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7408156531960746,
+      "learning_rate": 1.2210982964509837e-05,
+      "loss": 0.8389,
+      "step": 14727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.926698431416533,
+      "learning_rate": 1.2207399608415778e-05,
+      "loss": 1.2305,
+      "step": 14728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8278111236228294,
+      "learning_rate": 1.2203816608334572e-05,
+      "loss": 1.4102,
+      "step": 14729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8573275552155404,
+      "learning_rate": 1.2200233964365939e-05,
+      "loss": 1.1196,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.907339490081319,
+      "learning_rate": 1.2196651676609597e-05,
+      "loss": 1.7754,
+      "step": 14731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.617137169304874,
+      "learning_rate": 1.2193069745165223e-05,
+      "loss": 1.2559,
+      "step": 14732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9064593449762968,
+      "learning_rate": 1.2189488170132506e-05,
+      "loss": 0.9658,
+      "step": 14733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8999670954414911,
+      "learning_rate": 1.2185906951611105e-05,
+      "loss": 1.0615,
+      "step": 14734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9934480755398816,
+      "learning_rate": 1.2182326089700713e-05,
+      "loss": 0.6533,
+      "step": 14735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.619551772149005,
+      "learning_rate": 1.2178745584500964e-05,
+      "loss": 1.4619,
+      "step": 14736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8709117824812447,
+      "learning_rate": 1.2175165436111505e-05,
+      "loss": 0.8213,
+      "step": 14737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4514816179520869,
+      "learning_rate": 1.2171585644631964e-05,
+      "loss": 0.9504,
+      "step": 14738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3580497404939953,
+      "learning_rate": 1.216800621016197e-05,
+      "loss": 1.251,
+      "step": 14739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.364649053669073,
+      "learning_rate": 1.2164427132801148e-05,
+      "loss": 1.0752,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.011874539988163,
+      "learning_rate": 1.2160848412649095e-05,
+      "loss": 0.8477,
+      "step": 14741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.121334061901861,
+      "learning_rate": 1.2157270049805406e-05,
+      "loss": 0.9116,
+      "step": 14742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9208930383479346,
+      "learning_rate": 1.2153692044369649e-05,
+      "loss": 1.252,
+      "step": 14743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.35985080392871,
+      "learning_rate": 1.2150114396441425e-05,
+      "loss": 0.729,
+      "step": 14744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0716529618480317,
+      "learning_rate": 1.2146537106120287e-05,
+      "loss": 1.3711,
+      "step": 14745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9475298098984157,
+      "learning_rate": 1.2142960173505778e-05,
+      "loss": 1.25,
+      "step": 14746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.357456977149637,
+      "learning_rate": 1.2139383598697462e-05,
+      "loss": 1.207,
+      "step": 14747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.386754212061377,
+      "learning_rate": 1.2135807381794855e-05,
+      "loss": 0.5559,
+      "step": 14748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.122640917736673,
+      "learning_rate": 1.2132231522897505e-05,
+      "loss": 0.8613,
+      "step": 14749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5162639122127324,
+      "learning_rate": 1.212865602210491e-05,
+      "loss": 1.5,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.272139608452238,
+      "learning_rate": 1.2125080879516568e-05,
+      "loss": 1.4805,
+      "step": 14751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.130102203137912,
+      "learning_rate": 1.2121506095231994e-05,
+      "loss": 1.457,
+      "step": 14752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1532973930540455,
+      "learning_rate": 1.2117931669350662e-05,
+      "loss": 0.9216,
+      "step": 14753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.892316828329326,
+      "learning_rate": 1.2114357601972048e-05,
+      "loss": 1.3203,
+      "step": 14754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3600593231309905,
+      "learning_rate": 1.2110783893195607e-05,
+      "loss": 0.9395,
+      "step": 14755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.301313199527995,
+      "learning_rate": 1.2107210543120815e-05,
+      "loss": 0.9482,
+      "step": 14756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.748113157512205,
+      "learning_rate": 1.2103637551847091e-05,
+      "loss": 1.084,
+      "step": 14757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.249148419609928,
+      "learning_rate": 1.2100064919473894e-05,
+      "loss": 1.373,
+      "step": 14758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.169496510014273,
+      "learning_rate": 1.209649264610064e-05,
+      "loss": 1.1484,
+      "step": 14759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.987451146191286,
+      "learning_rate": 1.2092920731826734e-05,
+      "loss": 1.127,
+      "step": 14760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078913030143843,
+      "learning_rate": 1.2089349176751599e-05,
+      "loss": 0.7654,
+      "step": 14761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.962676937302407,
+      "learning_rate": 1.2085777980974619e-05,
+      "loss": 1.0615,
+      "step": 14762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2459885709870444,
+      "learning_rate": 1.2082207144595182e-05,
+      "loss": 0.8096,
+      "step": 14763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.866809788908865,
+      "learning_rate": 1.2078636667712649e-05,
+      "loss": 0.9688,
+      "step": 14764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.495472718829758,
+      "learning_rate": 1.2075066550426408e-05,
+      "loss": 1.0142,
+      "step": 14765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7679362066074513,
+      "learning_rate": 1.2071496792835795e-05,
+      "loss": 0.9021,
+      "step": 14766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.573931896568479,
+      "learning_rate": 1.2067927395040172e-05,
+      "loss": 0.9995,
+      "step": 14767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.641766331147053,
+      "learning_rate": 1.2064358357138855e-05,
+      "loss": 1.4316,
+      "step": 14768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7946404422695603,
+      "learning_rate": 1.2060789679231188e-05,
+      "loss": 1.0068,
+      "step": 14769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8380687274639933,
+      "learning_rate": 1.2057221361416476e-05,
+      "loss": 0.8464,
+      "step": 14770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.451167011384389,
+      "learning_rate": 1.2053653403794027e-05,
+      "loss": 1.4761,
+      "step": 14771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3118326409824976,
+      "learning_rate": 1.205008580646312e-05,
+      "loss": 0.7949,
+      "step": 14772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2317672758162246,
+      "learning_rate": 1.2046518569523066e-05,
+      "loss": 0.3535,
+      "step": 14773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.352895974432675,
+      "learning_rate": 1.2042951693073126e-05,
+      "loss": 1.4414,
+      "step": 14774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7610284844412103,
+      "learning_rate": 1.2039385177212556e-05,
+      "loss": 0.9041,
+      "step": 14775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.601212211602044,
+      "learning_rate": 1.2035819022040632e-05,
+      "loss": 1.0059,
+      "step": 14776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8997388233569383,
+      "learning_rate": 1.2032253227656572e-05,
+      "loss": 1.3633,
+      "step": 14777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3082969919766247,
+      "learning_rate": 1.2028687794159641e-05,
+      "loss": 1.2188,
+      "step": 14778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.08545967944359,
+      "learning_rate": 1.2025122721649043e-05,
+      "loss": 0.8657,
+      "step": 14779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8874476683795565,
+      "learning_rate": 1.2021558010224001e-05,
+      "loss": 1.3125,
+      "step": 14780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.613488075039498,
+      "learning_rate": 1.2017993659983707e-05,
+      "loss": 1.623,
+      "step": 14781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.326606780293634,
+      "learning_rate": 1.2014429671027369e-05,
+      "loss": 0.9482,
+      "step": 14782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.307549513164731,
+      "learning_rate": 1.2010866043454172e-05,
+      "loss": 0.8047,
+      "step": 14783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.402311793169055,
+      "learning_rate": 1.2007302777363283e-05,
+      "loss": 0.5219,
+      "step": 14784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.60066833616642,
+      "learning_rate": 1.200373987285385e-05,
+      "loss": 0.856,
+      "step": 14785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.508103820948025,
+      "learning_rate": 1.2000177330025064e-05,
+      "loss": 1.7852,
+      "step": 14786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.415130378390232,
+      "learning_rate": 1.199661514897605e-05,
+      "loss": 0.4888,
+      "step": 14787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2162903597750425,
+      "learning_rate": 1.1993053329805944e-05,
+      "loss": 1.1641,
+      "step": 14788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7521424184420287,
+      "learning_rate": 1.1989491872613859e-05,
+      "loss": 1.0391,
+      "step": 14789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.529568702350101,
+      "learning_rate": 1.1985930777498927e-05,
+      "loss": 0.8223,
+      "step": 14790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.214960727712326,
+      "learning_rate": 1.1982370044560245e-05,
+      "loss": 0.957,
+      "step": 14791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.102048194829041,
+      "learning_rate": 1.1978809673896905e-05,
+      "loss": 1.6289,
+      "step": 14792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.730101029898595,
+      "learning_rate": 1.197524966560798e-05,
+      "loss": 0.9224,
+      "step": 14793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5168331528677705,
+      "learning_rate": 1.1971690019792556e-05,
+      "loss": 1.1113,
+      "step": 14794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.072829275665237,
+      "learning_rate": 1.1968130736549704e-05,
+      "loss": 1.377,
+      "step": 14795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.9844897983449306,
+      "learning_rate": 1.1964571815978468e-05,
+      "loss": 1.0215,
+      "step": 14796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.002759362798238,
+      "learning_rate": 1.1961013258177892e-05,
+      "loss": 1.332,
+      "step": 14797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9488784839657867,
+      "learning_rate": 1.1957455063246997e-05,
+      "loss": 1.001,
+      "step": 14798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.742072295896232,
+      "learning_rate": 1.195389723128483e-05,
+      "loss": 1.4141,
+      "step": 14799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2429029007948156,
+      "learning_rate": 1.1950339762390391e-05,
+      "loss": 0.7412,
+      "step": 14800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4711524765846637,
+      "learning_rate": 1.1946782656662683e-05,
+      "loss": 1.0342,
+      "step": 14801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.703091747983475,
+      "learning_rate": 1.1943225914200693e-05,
+      "loss": 0.7905,
+      "step": 14802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5712139554758924,
+      "learning_rate": 1.1939669535103409e-05,
+      "loss": 0.5249,
+      "step": 14803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9317485644801846,
+      "learning_rate": 1.1936113519469814e-05,
+      "loss": 0.9512,
+      "step": 14804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0775178416995415,
+      "learning_rate": 1.1932557867398864e-05,
+      "loss": 0.6221,
+      "step": 14805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.900252448604928,
+      "learning_rate": 1.1929002578989496e-05,
+      "loss": 1.3125,
+      "step": 14806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.041197994085946,
+      "learning_rate": 1.1925447654340675e-05,
+      "loss": 0.8675,
+      "step": 14807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.552663934047614,
+      "learning_rate": 1.1921893093551322e-05,
+      "loss": 0.9678,
+      "step": 14808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.372352234844736,
+      "learning_rate": 1.1918338896720364e-05,
+      "loss": 0.6597,
+      "step": 14809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.655463569724109,
+      "learning_rate": 1.1914785063946699e-05,
+      "loss": 0.7852,
+      "step": 14810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0237104898862173,
+      "learning_rate": 1.1911231595329245e-05,
+      "loss": 0.8467,
+      "step": 14811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.746097425538228,
+      "learning_rate": 1.1907678490966879e-05,
+      "loss": 1.5352,
+      "step": 14812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3171935028869965,
+      "learning_rate": 1.19041257509585e-05,
+      "loss": 0.8789,
+      "step": 14813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8939115474272155,
+      "learning_rate": 1.1900573375402971e-05,
+      "loss": 1.0645,
+      "step": 14814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0144445230131325,
+      "learning_rate": 1.1897021364399141e-05,
+      "loss": 0.8445,
+      "step": 14815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9214079681379346,
+      "learning_rate": 1.1893469718045885e-05,
+      "loss": 1.2109,
+      "step": 14816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.968882507579415,
+      "learning_rate": 1.1889918436442027e-05,
+      "loss": 1.3672,
+      "step": 14817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.5085493744667025,
+      "learning_rate": 1.1886367519686406e-05,
+      "loss": 1.5781,
+      "step": 14818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1147535568891818,
+      "learning_rate": 1.1882816967877825e-05,
+      "loss": 0.7891,
+      "step": 14819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.533005282427027,
+      "learning_rate": 1.1879266781115119e-05,
+      "loss": 0.7944,
+      "step": 14820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.65015798853517,
+      "learning_rate": 1.1875716959497077e-05,
+      "loss": 1.4062,
+      "step": 14821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.725247916046989,
+      "learning_rate": 1.1872167503122481e-05,
+      "loss": 1.6367,
+      "step": 14822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1442041452052814,
+      "learning_rate": 1.1868618412090115e-05,
+      "loss": 0.9189,
+      "step": 14823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1657158770629064,
+      "learning_rate": 1.1865069686498767e-05,
+      "loss": 1.0742,
+      "step": 14824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4280701187981655,
+      "learning_rate": 1.1861521326447181e-05,
+      "loss": 0.916,
+      "step": 14825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3651604865303915,
+      "learning_rate": 1.1857973332034105e-05,
+      "loss": 0.7617,
+      "step": 14826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.70190854517529,
+      "learning_rate": 1.1854425703358275e-05,
+      "loss": 2.1621,
+      "step": 14827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.732978438563009,
+      "learning_rate": 1.1850878440518434e-05,
+      "loss": 1.3633,
+      "step": 14828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2312382769787042,
+      "learning_rate": 1.1847331543613293e-05,
+      "loss": 0.9897,
+      "step": 14829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.014468431074248,
+      "learning_rate": 1.184378501274156e-05,
+      "loss": 1.4043,
+      "step": 14830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.050500701606517,
+      "learning_rate": 1.1840238848001927e-05,
+      "loss": 0.7041,
+      "step": 14831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7156906453616787,
+      "learning_rate": 1.183669304949309e-05,
+      "loss": 0.9302,
+      "step": 14832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3464970222983115,
+      "learning_rate": 1.1833147617313731e-05,
+      "loss": 1.3799,
+      "step": 14833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.091509229107956,
+      "learning_rate": 1.1829602551562519e-05,
+      "loss": 0.7163,
+      "step": 14834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.566751720329389,
+      "learning_rate": 1.1826057852338105e-05,
+      "loss": 0.9824,
+      "step": 14835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.012419344664568,
+      "learning_rate": 1.1822513519739126e-05,
+      "loss": 1.0889,
+      "step": 14836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.056924688527764,
+      "learning_rate": 1.1818969553864243e-05,
+      "loss": 1.416,
+      "step": 14837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.325752183226568,
+      "learning_rate": 1.1815425954812071e-05,
+      "loss": 1.5039,
+      "step": 14838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5111291172522296,
+      "learning_rate": 1.1811882722681225e-05,
+      "loss": 0.7988,
+      "step": 14839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8915147815623907,
+      "learning_rate": 1.1808339857570305e-05,
+      "loss": 0.6865,
+      "step": 14840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.193266060377909,
+      "learning_rate": 1.1804797359577918e-05,
+      "loss": 1.4961,
+      "step": 14841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.395738983347032,
+      "learning_rate": 1.1801255228802661e-05,
+      "loss": 1.3086,
+      "step": 14842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.378395163470399,
+      "learning_rate": 1.1797713465343094e-05,
+      "loss": 1.2793,
+      "step": 14843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.437453113582941,
+      "learning_rate": 1.1794172069297776e-05,
+      "loss": 0.7866,
+      "step": 14844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3384621542232877,
+      "learning_rate": 1.1790631040765285e-05,
+      "loss": 1.0225,
+      "step": 14845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4792415283433296,
+      "learning_rate": 1.1787090379844153e-05,
+      "loss": 0.8242,
+      "step": 14846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7700487167084433,
+      "learning_rate": 1.1783550086632919e-05,
+      "loss": 0.7627,
+      "step": 14847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.890808537929942,
+      "learning_rate": 1.1780010161230095e-05,
+      "loss": 0.9534,
+      "step": 14848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.680800062082486,
+      "learning_rate": 1.1776470603734217e-05,
+      "loss": 1.1953,
+      "step": 14849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.683890735789916,
+      "learning_rate": 1.1772931414243769e-05,
+      "loss": 1.3018,
+      "step": 14850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.448948603429143,
+      "learning_rate": 1.1769392592857261e-05,
+      "loss": 1.2441,
+      "step": 14851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.131573624807627,
+      "learning_rate": 1.1765854139673174e-05,
+      "loss": 0.8486,
+      "step": 14852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.543046420902026,
+      "learning_rate": 1.1762316054789968e-05,
+      "loss": 0.7375,
+      "step": 14853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.058753454394242,
+      "learning_rate": 1.1758778338306128e-05,
+      "loss": 0.8574,
+      "step": 14854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.850982690679,
+      "learning_rate": 1.1755240990320097e-05,
+      "loss": 0.9893,
+      "step": 14855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.151184207219143,
+      "learning_rate": 1.1751704010930314e-05,
+      "loss": 0.9014,
+      "step": 14856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.899891186687825,
+      "learning_rate": 1.174816740023521e-05,
+      "loss": 1.1006,
+      "step": 14857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5694774093458106,
+      "learning_rate": 1.1744631158333218e-05,
+      "loss": 1.2654,
+      "step": 14858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6920482384722826,
+      "learning_rate": 1.1741095285322736e-05,
+      "loss": 1.0938,
+      "step": 14859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.414867226493168,
+      "learning_rate": 1.1737559781302185e-05,
+      "loss": 0.9258,
+      "step": 14860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.51875614381542,
+      "learning_rate": 1.1734024646369945e-05,
+      "loss": 1.291,
+      "step": 14861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.918549179972398,
+      "learning_rate": 1.173048988062439e-05,
+      "loss": 0.8926,
+      "step": 14862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.570620532840007,
+      "learning_rate": 1.1726955484163909e-05,
+      "loss": 1.1738,
+      "step": 14863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.808538147186262,
+      "learning_rate": 1.1723421457086855e-05,
+      "loss": 1.1162,
+      "step": 14864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5644414106858306,
+      "learning_rate": 1.1719887799491567e-05,
+      "loss": 1.0653,
+      "step": 14865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8740617630918104,
+      "learning_rate": 1.1716354511476401e-05,
+      "loss": 0.6587,
+      "step": 14866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7741854666472703,
+      "learning_rate": 1.1712821593139683e-05,
+      "loss": 1.5742,
+      "step": 14867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.7508524387235544,
+      "learning_rate": 1.1709289044579725e-05,
+      "loss": 0.9961,
+      "step": 14868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.430739785130267,
+      "learning_rate": 1.1705756865894847e-05,
+      "loss": 1.4414,
+      "step": 14869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.007570493652817,
+      "learning_rate": 1.1702225057183336e-05,
+      "loss": 0.8027,
+      "step": 14870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.525418664196865,
+      "learning_rate": 1.1698693618543497e-05,
+      "loss": 0.54,
+      "step": 14871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1940660353912445,
+      "learning_rate": 1.16951625500736e-05,
+      "loss": 1.498,
+      "step": 14872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2945801565733515,
+      "learning_rate": 1.1691631851871912e-05,
+      "loss": 0.8799,
+      "step": 14873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4657716787505806,
+      "learning_rate": 1.1688101524036682e-05,
+      "loss": 0.6289,
+      "step": 14874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7166688368809497,
+      "learning_rate": 1.1684571566666177e-05,
+      "loss": 1.4092,
+      "step": 14875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6084658315763236,
+      "learning_rate": 1.1681041979858626e-05,
+      "loss": 0.8989,
+      "step": 14876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.604913964477432,
+      "learning_rate": 1.1677512763712245e-05,
+      "loss": 0.7607,
+      "step": 14877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.0955557470769355,
+      "learning_rate": 1.1673983918325268e-05,
+      "loss": 1.4805,
+      "step": 14878
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3190897047535013,
+      "learning_rate": 1.1670455443795885e-05,
+      "loss": 0.9062,
+      "step": 14879
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7356030348358096,
+      "learning_rate": 1.1666927340222312e-05,
+      "loss": 1.3828,
+      "step": 14880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.141185046312653,
+      "learning_rate": 1.166339960770272e-05,
+      "loss": 1.3887,
+      "step": 14881
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.446160994963588,
+      "learning_rate": 1.165987224633529e-05,
+      "loss": 1.3555,
+      "step": 14882
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.260672912104428,
+      "learning_rate": 1.1656345256218173e-05,
+      "loss": 1.0352,
+      "step": 14883
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4515247348085336,
+      "learning_rate": 1.1652818637449544e-05,
+      "loss": 0.8137,
+      "step": 14884
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.239323796037202,
+      "learning_rate": 1.1649292390127541e-05,
+      "loss": 1.3066,
+      "step": 14885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.682004033373976,
+      "learning_rate": 1.1645766514350288e-05,
+      "loss": 1.076,
+      "step": 14886
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.961889988306226,
+      "learning_rate": 1.1642241010215913e-05,
+      "loss": 0.9116,
+      "step": 14887
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2052831157485127,
+      "learning_rate": 1.1638715877822543e-05,
+      "loss": 1.1611,
+      "step": 14888
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4310118570858497,
+      "learning_rate": 1.1635191117268271e-05,
+      "loss": 0.998,
+      "step": 14889
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.75352477676875,
+      "learning_rate": 1.1631666728651188e-05,
+      "loss": 1.2793,
+      "step": 14890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.351479411484646,
+      "learning_rate": 1.1628142712069373e-05,
+      "loss": 1.2703,
+      "step": 14891
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2454997964146957,
+      "learning_rate": 1.1624619067620906e-05,
+      "loss": 1.0137,
+      "step": 14892
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.062185668112075,
+      "learning_rate": 1.1621095795403849e-05,
+      "loss": 1.1348,
+      "step": 14893
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.870134053484095,
+      "learning_rate": 1.161757289551625e-05,
+      "loss": 1.3379,
+      "step": 14894
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.580144339206871,
+      "learning_rate": 1.1614050368056137e-05,
+      "loss": 0.751,
+      "step": 14895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6141106126260123,
+      "learning_rate": 1.1610528213121555e-05,
+      "loss": 1.2285,
+      "step": 14896
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8974469901459219,
+      "learning_rate": 1.1607006430810529e-05,
+      "loss": 1.4062,
+      "step": 14897
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.119165774869067,
+      "learning_rate": 1.1603485021221064e-05,
+      "loss": 1.4121,
+      "step": 14898
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.412640100077598,
+      "learning_rate": 1.1599963984451156e-05,
+      "loss": 0.8473,
+      "step": 14899
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3254829110317563,
+      "learning_rate": 1.1596443320598788e-05,
+      "loss": 0.9932,
+      "step": 14900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.344764122464178,
+      "learning_rate": 1.1592923029761951e-05,
+      "loss": 1.4316,
+      "step": 14901
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.975704625138294,
+      "learning_rate": 1.1589403112038613e-05,
+      "loss": 1.2421,
+      "step": 14902
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.020639813498306,
+      "learning_rate": 1.1585883567526715e-05,
+      "loss": 0.835,
+      "step": 14903
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8759931706702,
+      "learning_rate": 1.1582364396324225e-05,
+      "loss": 1.0723,
+      "step": 14904
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.429057735751059,
+      "learning_rate": 1.1578845598529062e-05,
+      "loss": 0.4038,
+      "step": 14905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.610439785317458,
+      "learning_rate": 1.1575327174239173e-05,
+      "loss": 1.4727,
+      "step": 14906
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.029309655972216,
+      "learning_rate": 1.1571809123552462e-05,
+      "loss": 1.0986,
+      "step": 14907
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.621011072953573,
+      "learning_rate": 1.1568291446566828e-05,
+      "loss": 2.2812,
+      "step": 14908
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.002569931657009,
+      "learning_rate": 1.1564774143380186e-05,
+      "loss": 0.7725,
+      "step": 14909
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.303535985885505,
+      "learning_rate": 1.156125721409041e-05,
+      "loss": 0.688,
+      "step": 14910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.5316744901517785,
+      "learning_rate": 1.1557740658795371e-05,
+      "loss": 0.6006,
+      "step": 14911
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.02323455163282,
+      "learning_rate": 1.1554224477592932e-05,
+      "loss": 1.1865,
+      "step": 14912
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.047538020058751,
+      "learning_rate": 1.155070867058096e-05,
+      "loss": 0.9551,
+      "step": 14913
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6163834075634855,
+      "learning_rate": 1.154719323785728e-05,
+      "loss": 1.1348,
+      "step": 14914
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.016939919316599,
+      "learning_rate": 1.1543678179519745e-05,
+      "loss": 0.9365,
+      "step": 14915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.602331778564946,
+      "learning_rate": 1.154016349566617e-05,
+      "loss": 0.6758,
+      "step": 14916
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.634489074626458,
+      "learning_rate": 1.1536649186394355e-05,
+      "loss": 1.1846,
+      "step": 14917
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.985201861359864,
+      "learning_rate": 1.1533135251802121e-05,
+      "loss": 1.5752,
+      "step": 14918
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.2001213132939155,
+      "learning_rate": 1.1529621691987252e-05,
+      "loss": 0.7908,
+      "step": 14919
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.567935579824411,
+      "learning_rate": 1.1526108507047525e-05,
+      "loss": 0.877,
+      "step": 14920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.346840918403965,
+      "learning_rate": 1.1522595697080707e-05,
+      "loss": 0.6201,
+      "step": 14921
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.523865828471673,
+      "learning_rate": 1.1519083262184571e-05,
+      "loss": 0.8701,
+      "step": 14922
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.192252113558476,
+      "learning_rate": 1.1515571202456859e-05,
+      "loss": 1.3105,
+      "step": 14923
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.752877204126798,
+      "learning_rate": 1.1512059517995305e-05,
+      "loss": 0.7207,
+      "step": 14924
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.34950040381851,
+      "learning_rate": 1.150854820889764e-05,
+      "loss": 1.25,
+      "step": 14925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.636597615327636,
+      "learning_rate": 1.1505037275261598e-05,
+      "loss": 1.4863,
+      "step": 14926
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8566105270915736,
+      "learning_rate": 1.1501526717184874e-05,
+      "loss": 1.0454,
+      "step": 14927
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.130035531737063,
+      "learning_rate": 1.1498016534765165e-05,
+      "loss": 1.1816,
+      "step": 14928
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1974398802430786,
+      "learning_rate": 1.1494506728100151e-05,
+      "loss": 0.9551,
+      "step": 14929
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.901924705073404,
+      "learning_rate": 1.1490997297287525e-05,
+      "loss": 1.1504,
+      "step": 14930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7498246469825327,
+      "learning_rate": 1.1487488242424948e-05,
+      "loss": 1.0586,
+      "step": 14931
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3299259670103827,
+      "learning_rate": 1.148397956361007e-05,
+      "loss": 1.6387,
+      "step": 14932
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.026074703549399,
+      "learning_rate": 1.148047126094053e-05,
+      "loss": 1.2383,
+      "step": 14933
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.412036527081896,
+      "learning_rate": 1.1476963334513967e-05,
+      "loss": 1.5488,
+      "step": 14934
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.74329243942365,
+      "learning_rate": 1.1473455784428022e-05,
+      "loss": 0.6633,
+      "step": 14935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0372173915679097,
+      "learning_rate": 1.1469948610780296e-05,
+      "loss": 0.8975,
+      "step": 14936
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.570333755880704,
+      "learning_rate": 1.146644181366839e-05,
+      "loss": 0.6115,
+      "step": 14937
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.477706789558883,
+      "learning_rate": 1.1462935393189891e-05,
+      "loss": 0.9727,
+      "step": 14938
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.29712018207595,
+      "learning_rate": 1.1459429349442399e-05,
+      "loss": 1.75,
+      "step": 14939
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.307440828343007,
+      "learning_rate": 1.1455923682523475e-05,
+      "loss": 1.1309,
+      "step": 14940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.218785942772235,
+      "learning_rate": 1.1452418392530682e-05,
+      "loss": 1.2705,
+      "step": 14941
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4608671632433543,
+      "learning_rate": 1.1448913479561552e-05,
+      "loss": 1.248,
+      "step": 14942
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.127263826738424,
+      "learning_rate": 1.144540894371366e-05,
+      "loss": 0.7395,
+      "step": 14943
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6070723866347119,
+      "learning_rate": 1.1441904785084517e-05,
+      "loss": 0.4954,
+      "step": 14944
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4763039053798135,
+      "learning_rate": 1.1438401003771648e-05,
+      "loss": 1.1074,
+      "step": 14945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4758852444803354,
+      "learning_rate": 1.1434897599872546e-05,
+      "loss": 1.3525,
+      "step": 14946
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.569432070132406,
+      "learning_rate": 1.1431394573484733e-05,
+      "loss": 1.3301,
+      "step": 14947
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.140521147420055,
+      "learning_rate": 1.1427891924705685e-05,
+      "loss": 0.9883,
+      "step": 14948
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3569550893564792,
+      "learning_rate": 1.1424389653632881e-05,
+      "loss": 0.8691,
+      "step": 14949
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8918278021719326,
+      "learning_rate": 1.1420887760363774e-05,
+      "loss": 1.2676,
+      "step": 14950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.694647046192181,
+      "learning_rate": 1.1417386244995832e-05,
+      "loss": 1.2168,
+      "step": 14951
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.475353656583755,
+      "learning_rate": 1.1413885107626512e-05,
+      "loss": 0.7002,
+      "step": 14952
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.285251841935431,
+      "learning_rate": 1.141038434835324e-05,
+      "loss": 0.9863,
+      "step": 14953
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.8713780177177695,
+      "learning_rate": 1.1406883967273437e-05,
+      "loss": 1.1333,
+      "step": 14954
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.366000441234351,
+      "learning_rate": 1.1403383964484513e-05,
+      "loss": 1.1637,
+      "step": 14955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1534653526124687,
+      "learning_rate": 1.1399884340083886e-05,
+      "loss": 0.8916,
+      "step": 14956
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.249194310445507,
+      "learning_rate": 1.1396385094168941e-05,
+      "loss": 1.3105,
+      "step": 14957
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.707818691556954,
+      "learning_rate": 1.1392886226837057e-05,
+      "loss": 1.4238,
+      "step": 14958
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.350107494209958,
+      "learning_rate": 1.1389387738185606e-05,
+      "loss": 1.1904,
+      "step": 14959
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.1249905210443885,
+      "learning_rate": 1.1385889628311961e-05,
+      "loss": 1.1162,
+      "step": 14960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.047802539683234,
+      "learning_rate": 1.1382391897313455e-05,
+      "loss": 1.0581,
+      "step": 14961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2127035790874134,
+      "learning_rate": 1.137889454528745e-05,
+      "loss": 1.1914,
+      "step": 14962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.186125516080367,
+      "learning_rate": 1.1375397572331253e-05,
+      "loss": 1.8848,
+      "step": 14963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2901586830764926,
+      "learning_rate": 1.1371900978542206e-05,
+      "loss": 1.3301,
+      "step": 14964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.64531773827491,
+      "learning_rate": 1.1368404764017607e-05,
+      "loss": 1.0166,
+      "step": 14965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9285969650005863,
+      "learning_rate": 1.1364908928854753e-05,
+      "loss": 1.1445,
+      "step": 14966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.571205091122845,
+      "learning_rate": 1.1361413473150922e-05,
+      "loss": 0.7852,
+      "step": 14967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.745387737742285,
+      "learning_rate": 1.1357918397003413e-05,
+      "loss": 0.6765,
+      "step": 14968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0464983877571292,
+      "learning_rate": 1.1354423700509482e-05,
+      "loss": 1.5879,
+      "step": 14969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2795362084156916,
+      "learning_rate": 1.1350929383766374e-05,
+      "loss": 1.4551,
+      "step": 14970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.229774936187858,
+      "learning_rate": 1.1347435446871354e-05,
+      "loss": 0.9258,
+      "step": 14971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.699811551649053,
+      "learning_rate": 1.134394188992164e-05,
+      "loss": 1.1719,
+      "step": 14972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.720472908001505,
+      "learning_rate": 1.134044871301447e-05,
+      "loss": 1.0728,
+      "step": 14973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4242400404167013,
+      "learning_rate": 1.1336955916247052e-05,
+      "loss": 0.7949,
+      "step": 14974
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3067582477551536,
+      "learning_rate": 1.133346349971659e-05,
+      "loss": 0.9351,
+      "step": 14975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3532400490113954,
+      "learning_rate": 1.1329971463520265e-05,
+      "loss": 1.0869,
+      "step": 14976
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.922070259005714,
+      "learning_rate": 1.1326479807755277e-05,
+      "loss": 0.9053,
+      "step": 14977
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8751288051867956,
+      "learning_rate": 1.1322988532518794e-05,
+      "loss": 0.9463,
+      "step": 14978
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.480525459181966,
+      "learning_rate": 1.1319497637907959e-05,
+      "loss": 1.3184,
+      "step": 14979
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.888081036778885,
+      "learning_rate": 1.1316007124019937e-05,
+      "loss": 0.9775,
+      "step": 14980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.50551511878037,
+      "learning_rate": 1.1312516990951874e-05,
+      "loss": 1.0911,
+      "step": 14981
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4076141850061386,
+      "learning_rate": 1.1309027238800892e-05,
+      "loss": 0.6807,
+      "step": 14982
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0119844114145895,
+      "learning_rate": 1.130553786766411e-05,
+      "loss": 1.0156,
+      "step": 14983
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.143101565034149,
+      "learning_rate": 1.1302048877638622e-05,
+      "loss": 1.4199,
+      "step": 14984
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.4078281903993375,
+      "learning_rate": 1.1298560268821548e-05,
+      "loss": 1.1543,
+      "step": 14985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.025358959454431,
+      "learning_rate": 1.1295072041309963e-05,
+      "loss": 0.8267,
+      "step": 14986
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5538706081881557,
+      "learning_rate": 1.1291584195200943e-05,
+      "loss": 0.9688,
+      "step": 14987
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.087388623460615,
+      "learning_rate": 1.1288096730591547e-05,
+      "loss": 1.4004,
+      "step": 14988
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2962404788945916,
+      "learning_rate": 1.1284609647578836e-05,
+      "loss": 0.9395,
+      "step": 14989
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5970451840893922,
+      "learning_rate": 1.1281122946259865e-05,
+      "loss": 1.3203,
+      "step": 14990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2991058518912704,
+      "learning_rate": 1.1277636626731655e-05,
+      "loss": 0.7422,
+      "step": 14991
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.590882113263295,
+      "learning_rate": 1.1274150689091234e-05,
+      "loss": 1.5,
+      "step": 14992
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.381871171966736,
+      "learning_rate": 1.1270665133435602e-05,
+      "loss": 0.8145,
+      "step": 14993
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.306856559867843,
+      "learning_rate": 1.1267179959861777e-05,
+      "loss": 1.4639,
+      "step": 14994
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7809153151856987,
+      "learning_rate": 1.1263695168466745e-05,
+      "loss": 1.4395,
+      "step": 14995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.694319989863955,
+      "learning_rate": 1.1260210759347483e-05,
+      "loss": 1.0986,
+      "step": 14996
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.332534531536509,
+      "learning_rate": 1.1256726732600956e-05,
+      "loss": 1.2832,
+      "step": 14997
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.75501176004271,
+      "learning_rate": 1.1253243088324125e-05,
+      "loss": 1.1025,
+      "step": 14998
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.420035507954762,
+      "learning_rate": 1.1249759826613954e-05,
+      "loss": 1.043,
+      "step": 14999
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5961530184279622,
+      "learning_rate": 1.124627694756737e-05,
+      "loss": 0.8843,
+      "step": 15000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.175568797231656,
+      "learning_rate": 1.1242794451281289e-05,
+      "loss": 1.5693,
+      "step": 15001
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.370547990661968,
+      "learning_rate": 1.1239312337852647e-05,
+      "loss": 0.9307,
+      "step": 15002
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8694796430068745,
+      "learning_rate": 1.1235830607378341e-05,
+      "loss": 0.8066,
+      "step": 15003
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.785935601201417,
+      "learning_rate": 1.1232349259955265e-05,
+      "loss": 1.0781,
+      "step": 15004
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1578812993577565,
+      "learning_rate": 1.1228868295680297e-05,
+      "loss": 1.1602,
+      "step": 15005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1109310486057504,
+      "learning_rate": 1.1225387714650326e-05,
+      "loss": 0.9531,
+      "step": 15006
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.66881182730512,
+      "learning_rate": 1.12219075169622e-05,
+      "loss": 0.7827,
+      "step": 15007
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8518826493598675,
+      "learning_rate": 1.1218427702712787e-05,
+      "loss": 0.501,
+      "step": 15008
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5616304736619897,
+      "learning_rate": 1.1214948271998918e-05,
+      "loss": 1.3398,
+      "step": 15009
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.999853894960058,
+      "learning_rate": 1.121146922491742e-05,
+      "loss": 1.1769,
+      "step": 15010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5286204490343462,
+      "learning_rate": 1.1207990561565129e-05,
+      "loss": 0.7358,
+      "step": 15011
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.644205498741347,
+      "learning_rate": 1.1204512282038846e-05,
+      "loss": 0.9502,
+      "step": 15012
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0858151182220324,
+      "learning_rate": 1.1201034386435368e-05,
+      "loss": 0.7725,
+      "step": 15013
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.722852607664063,
+      "learning_rate": 1.1197556874851478e-05,
+      "loss": 0.8232,
+      "step": 15014
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8550485701028774,
+      "learning_rate": 1.119407974738397e-05,
+      "loss": 0.9473,
+      "step": 15015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9426612405123014,
+      "learning_rate": 1.1190603004129593e-05,
+      "loss": 0.9839,
+      "step": 15016
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0035641479144752,
+      "learning_rate": 1.1187126645185122e-05,
+      "loss": 1.0596,
+      "step": 15017
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.654578816399009,
+      "learning_rate": 1.1183650670647295e-05,
+      "loss": 0.9502,
+      "step": 15018
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.439260274840819,
+      "learning_rate": 1.1180175080612834e-05,
+      "loss": 0.7153,
+      "step": 15019
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.541070700510565,
+      "learning_rate": 1.1176699875178485e-05,
+      "loss": 0.8345,
+      "step": 15020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.406718419047809,
+      "learning_rate": 1.117322505444095e-05,
+      "loss": 1.0312,
+      "step": 15021
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5005308541307216,
+      "learning_rate": 1.1169750618496925e-05,
+      "loss": 0.7583,
+      "step": 15022
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3393201544669373,
+      "learning_rate": 1.1166276567443115e-05,
+      "loss": 0.9619,
+      "step": 15023
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1098234477009563,
+      "learning_rate": 1.11628029013762e-05,
+      "loss": 0.9893,
+      "step": 15024
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.931265680033113,
+      "learning_rate": 1.1159329620392847e-05,
+      "loss": 1.3428,
+      "step": 15025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8363552994260997,
+      "learning_rate": 1.1155856724589706e-05,
+      "loss": 0.6812,
+      "step": 15026
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0943156023975984,
+      "learning_rate": 1.1152384214063438e-05,
+      "loss": 0.7725,
+      "step": 15027
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5666855434262823,
+      "learning_rate": 1.1148912088910687e-05,
+      "loss": 0.9482,
+      "step": 15028
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.686261890556481,
+      "learning_rate": 1.1145440349228075e-05,
+      "loss": 1.5039,
+      "step": 15029
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6357098945259145,
+      "learning_rate": 1.1141968995112218e-05,
+      "loss": 0.8911,
+      "step": 15030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.351938920927722,
+      "learning_rate": 1.1138498026659711e-05,
+      "loss": 1.1484,
+      "step": 15031
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.247851087328345,
+      "learning_rate": 1.1135027443967173e-05,
+      "loss": 1.4844,
+      "step": 15032
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.170905353660389,
+      "learning_rate": 1.1131557247131174e-05,
+      "loss": 1.5332,
+      "step": 15033
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3748124701096875,
+      "learning_rate": 1.1128087436248294e-05,
+      "loss": 0.8994,
+      "step": 15034
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6725676187687424,
+      "learning_rate": 1.1124618011415083e-05,
+      "loss": 1.0557,
+      "step": 15035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8911995133079833,
+      "learning_rate": 1.1121148972728104e-05,
+      "loss": 0.6309,
+      "step": 15036
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9392675520365823,
+      "learning_rate": 1.1117680320283907e-05,
+      "loss": 1.0488,
+      "step": 15037
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.942620354111315,
+      "learning_rate": 1.1114212054179016e-05,
+      "loss": 0.8267,
+      "step": 15038
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.481613364851768,
+      "learning_rate": 1.1110744174509952e-05,
+      "loss": 0.9712,
+      "step": 15039
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1995526001021317,
+      "learning_rate": 1.1107276681373214e-05,
+      "loss": 1.4199,
+      "step": 15040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.060958335033364,
+      "learning_rate": 1.1103809574865318e-05,
+      "loss": 0.5493,
+      "step": 15041
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.550448385721852,
+      "learning_rate": 1.1100342855082743e-05,
+      "loss": 0.7363,
+      "step": 15042
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1774581391858017,
+      "learning_rate": 1.109687652212196e-05,
+      "loss": 1.1484,
+      "step": 15043
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.632901062876646,
+      "learning_rate": 1.1093410576079444e-05,
+      "loss": 1.1807,
+      "step": 15044
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1693748519400655,
+      "learning_rate": 1.108994501705166e-05,
+      "loss": 1.2715,
+      "step": 15045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0671023258299672,
+      "learning_rate": 1.1086479845135042e-05,
+      "loss": 0.8169,
+      "step": 15046
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.627102509402644,
+      "learning_rate": 1.1083015060426028e-05,
+      "loss": 0.7725,
+      "step": 15047
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4630303868352126,
+      "learning_rate": 1.1079550663021026e-05,
+      "loss": 1.3613,
+      "step": 15048
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4201471871942677,
+      "learning_rate": 1.1076086653016473e-05,
+      "loss": 1.3867,
+      "step": 15049
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7676134450485277,
+      "learning_rate": 1.1072623030508764e-05,
+      "loss": 1.0459,
+      "step": 15050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6136598460861435,
+      "learning_rate": 1.106915979559428e-05,
+      "loss": 0.709,
+      "step": 15051
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2427038310152767,
+      "learning_rate": 1.1065696948369403e-05,
+      "loss": 1.0819,
+      "step": 15052
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.210647827829452,
+      "learning_rate": 1.1062234488930506e-05,
+      "loss": 1.2236,
+      "step": 15053
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.811112573535022,
+      "learning_rate": 1.1058772417373958e-05,
+      "loss": 0.6787,
+      "step": 15054
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.388836085493552,
+      "learning_rate": 1.1055310733796099e-05,
+      "loss": 1.0596,
+      "step": 15055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9877586052552039,
+      "learning_rate": 1.1051849438293264e-05,
+      "loss": 0.7312,
+      "step": 15056
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.464512449773723,
+      "learning_rate": 1.104838853096177e-05,
+      "loss": 0.7961,
+      "step": 15057
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.418336132906776,
+      "learning_rate": 1.1044928011897957e-05,
+      "loss": 1.1162,
+      "step": 15058
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.718681398929024,
+      "learning_rate": 1.1041467881198112e-05,
+      "loss": 1.1855,
+      "step": 15059
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3605669055160203,
+      "learning_rate": 1.1038008138958528e-05,
+      "loss": 1.5723,
+      "step": 15060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.23234594020708,
+      "learning_rate": 1.1034548785275501e-05,
+      "loss": 1.6113,
+      "step": 15061
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.627773381282392,
+      "learning_rate": 1.1031089820245293e-05,
+      "loss": 1.0693,
+      "step": 15062
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9944214506367528,
+      "learning_rate": 1.1027631243964165e-05,
+      "loss": 0.8096,
+      "step": 15063
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0811648972899564,
+      "learning_rate": 1.1024173056528378e-05,
+      "loss": 0.6729,
+      "step": 15064
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3353159412533193,
+      "learning_rate": 1.1020715258034153e-05,
+      "loss": 1.1689,
+      "step": 15065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2277095583931934,
+      "learning_rate": 1.1017257848577745e-05,
+      "loss": 1.2812,
+      "step": 15066
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3490293933300963,
+      "learning_rate": 1.1013800828255358e-05,
+      "loss": 1.0547,
+      "step": 15067
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2184221415528746,
+      "learning_rate": 1.1010344197163199e-05,
+      "loss": 1.543,
+      "step": 15068
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5381384984365227,
+      "learning_rate": 1.1006887955397458e-05,
+      "loss": 0.9863,
+      "step": 15069
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.322749466324459,
+      "learning_rate": 1.1003432103054334e-05,
+      "loss": 1.2012,
+      "step": 15070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3584226997975595,
+      "learning_rate": 1.099997664023e-05,
+      "loss": 1.5059,
+      "step": 15071
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.02857809227825,
+      "learning_rate": 1.0996521567020607e-05,
+      "loss": 1.7617,
+      "step": 15072
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.467001573491764,
+      "learning_rate": 1.0993066883522327e-05,
+      "loss": 1.3555,
+      "step": 15073
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4526377394452137,
+      "learning_rate": 1.0989612589831286e-05,
+      "loss": 0.6943,
+      "step": 15074
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.837659660302416,
+      "learning_rate": 1.098615868604363e-05,
+      "loss": 1.6855,
+      "step": 15075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2422808897976694,
+      "learning_rate": 1.0982705172255472e-05,
+      "loss": 0.6353,
+      "step": 15076
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.943323244305306,
+      "learning_rate": 1.0979252048562923e-05,
+      "loss": 1.2383,
+      "step": 15077
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.548141807641564,
+      "learning_rate": 1.0975799315062075e-05,
+      "loss": 0.8512,
+      "step": 15078
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.607967677387686,
+      "learning_rate": 1.097234697184903e-05,
+      "loss": 1.2061,
+      "step": 15079
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.359317379724066,
+      "learning_rate": 1.096889501901986e-05,
+      "loss": 0.582,
+      "step": 15080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.1702587411501995,
+      "learning_rate": 1.0965443456670615e-05,
+      "loss": 1.918,
+      "step": 15081
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8811476947884103,
+      "learning_rate": 1.0961992284897371e-05,
+      "loss": 0.9556,
+      "step": 15082
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.732477703704978,
+      "learning_rate": 1.0958541503796173e-05,
+      "loss": 0.6702,
+      "step": 15083
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2977453126997545,
+      "learning_rate": 1.0955091113463052e-05,
+      "loss": 0.9443,
+      "step": 15084
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.023113696676767,
+      "learning_rate": 1.0951641113994023e-05,
+      "loss": 0.7139,
+      "step": 15085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.12515807966276,
+      "learning_rate": 1.0948191505485095e-05,
+      "loss": 0.853,
+      "step": 15086
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.074225328072623,
+      "learning_rate": 1.0944742288032287e-05,
+      "loss": 1.3809,
+      "step": 15087
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5017003969737353,
+      "learning_rate": 1.094129346173158e-05,
+      "loss": 1.1309,
+      "step": 15088
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.360344467443832,
+      "learning_rate": 1.093784502667895e-05,
+      "loss": 0.7788,
+      "step": 15089
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.079620855511755,
+      "learning_rate": 1.0934396982970364e-05,
+      "loss": 0.667,
+      "step": 15090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9150348918116555,
+      "learning_rate": 1.093094933070178e-05,
+      "loss": 1.3086,
+      "step": 15091
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.236277292666198,
+      "learning_rate": 1.0927502069969155e-05,
+      "loss": 1.1533,
+      "step": 15092
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.589430493632064,
+      "learning_rate": 1.0924055200868425e-05,
+      "loss": 0.9873,
+      "step": 15093
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.236634932215654,
+      "learning_rate": 1.0920608723495505e-05,
+      "loss": 1.1152,
+      "step": 15094
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8615482996664274,
+      "learning_rate": 1.0917162637946302e-05,
+      "loss": 1.0195,
+      "step": 15095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6669501859695117,
+      "learning_rate": 1.091371694431674e-05,
+      "loss": 1.1211,
+      "step": 15096
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.832268477402589,
+      "learning_rate": 1.0910271642702702e-05,
+      "loss": 1.0039,
+      "step": 15097
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7874538506947615,
+      "learning_rate": 1.0906826733200067e-05,
+      "loss": 1.1426,
+      "step": 15098
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6324451767591626,
+      "learning_rate": 1.0903382215904697e-05,
+      "loss": 1.666,
+      "step": 15099
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.095561692327517,
+      "learning_rate": 1.0899938090912464e-05,
+      "loss": 1.332,
+      "step": 15100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.638887741687706,
+      "learning_rate": 1.089649435831922e-05,
+      "loss": 1.1445,
+      "step": 15101
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.893585668392195,
+      "learning_rate": 1.0893051018220799e-05,
+      "loss": 1.3164,
+      "step": 15102
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.217447785254113,
+      "learning_rate": 1.0889608070713014e-05,
+      "loss": 0.4976,
+      "step": 15103
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6649985389588506,
+      "learning_rate": 1.0886165515891703e-05,
+      "loss": 0.4377,
+      "step": 15104
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4980357082615443,
+      "learning_rate": 1.0882723353852658e-05,
+      "loss": 0.8915,
+      "step": 15105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.324347792176675,
+      "learning_rate": 1.0879281584691676e-05,
+      "loss": 0.6604,
+      "step": 15106
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4764378272604337,
+      "learning_rate": 1.0875840208504529e-05,
+      "loss": 0.8457,
+      "step": 15107
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9707975052808315,
+      "learning_rate": 1.0872399225387009e-05,
+      "loss": 0.8003,
+      "step": 15108
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1939978956701016,
+      "learning_rate": 1.086895863543486e-05,
+      "loss": 1.3008,
+      "step": 15109
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.035697528225183,
+      "learning_rate": 1.0865518438743843e-05,
+      "loss": 0.9875,
+      "step": 15110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.759942807005811,
+      "learning_rate": 1.0862078635409697e-05,
+      "loss": 0.7617,
+      "step": 15111
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.955777275804089,
+      "learning_rate": 1.085863922552814e-05,
+      "loss": 1.1055,
+      "step": 15112
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.055806250649599,
+      "learning_rate": 1.0855200209194904e-05,
+      "loss": 1.416,
+      "step": 15113
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3449389759193684,
+      "learning_rate": 1.0851761586505687e-05,
+      "loss": 0.7593,
+      "step": 15114
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.409953407508574,
+      "learning_rate": 1.0848323357556187e-05,
+      "loss": 0.7961,
+      "step": 15115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.432335885168076,
+      "learning_rate": 1.0844885522442074e-05,
+      "loss": 0.8184,
+      "step": 15116
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.664646476214187,
+      "learning_rate": 1.0841448081259048e-05,
+      "loss": 1.3262,
+      "step": 15117
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.792526583980138,
+      "learning_rate": 1.0838011034102746e-05,
+      "loss": 1.2188,
+      "step": 15118
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.673449925540648,
+      "learning_rate": 1.083457438106884e-05,
+      "loss": 0.9307,
+      "step": 15119
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1179589699435297,
+      "learning_rate": 1.0831138122252954e-05,
+      "loss": 0.5166,
+      "step": 15120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.356885860273216,
+      "learning_rate": 1.0827702257750739e-05,
+      "loss": 1.2969,
+      "step": 15121
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.596880941923576,
+      "learning_rate": 1.0824266787657794e-05,
+      "loss": 0.7969,
+      "step": 15122
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.375839806743641,
+      "learning_rate": 1.0820831712069738e-05,
+      "loss": 1.1621,
+      "step": 15123
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9308452762349102,
+      "learning_rate": 1.0817397031082153e-05,
+      "loss": 1.5195,
+      "step": 15124
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.473390242048764,
+      "learning_rate": 1.0813962744790642e-05,
+      "loss": 0.7031,
+      "step": 15125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.479337463450522,
+      "learning_rate": 1.0810528853290775e-05,
+      "loss": 1.0596,
+      "step": 15126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2011677590495013,
+      "learning_rate": 1.0807095356678112e-05,
+      "loss": 0.8242,
+      "step": 15127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6946252803704134,
+      "learning_rate": 1.0803662255048199e-05,
+      "loss": 0.7344,
+      "step": 15128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.262968598252388,
+      "learning_rate": 1.0800229548496584e-05,
+      "loss": 0.8955,
+      "step": 15129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.277623049848712,
+      "learning_rate": 1.0796797237118813e-05,
+      "loss": 0.9189,
+      "step": 15130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.862606286620402,
+      "learning_rate": 1.079336532101039e-05,
+      "loss": 0.6211,
+      "step": 15131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.765559152367742,
+      "learning_rate": 1.0789933800266827e-05,
+      "loss": 1.1289,
+      "step": 15132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1543558619318417,
+      "learning_rate": 1.0786502674983615e-05,
+      "loss": 0.9551,
+      "step": 15133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.64222831130585,
+      "learning_rate": 1.0783071945256254e-05,
+      "loss": 0.3203,
+      "step": 15134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3024837279225707,
+      "learning_rate": 1.0779641611180211e-05,
+      "loss": 0.9355,
+      "step": 15135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7930192849449176,
+      "learning_rate": 1.0776211672850956e-05,
+      "loss": 1.5273,
+      "step": 15136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9178941638374212,
+      "learning_rate": 1.077278213036392e-05,
+      "loss": 0.8965,
+      "step": 15137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5320542553095216,
+      "learning_rate": 1.0769352983814585e-05,
+      "loss": 0.7695,
+      "step": 15138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7106226317729814,
+      "learning_rate": 1.0765924233298361e-05,
+      "loss": 1.3672,
+      "step": 15139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.282509445019388,
+      "learning_rate": 1.0762495878910672e-05,
+      "loss": 1.1985,
+      "step": 15140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.012770809173575,
+      "learning_rate": 1.0759067920746918e-05,
+      "loss": 1.0156,
+      "step": 15141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5113975117767007,
+      "learning_rate": 1.0755640358902514e-05,
+      "loss": 1.3594,
+      "step": 15142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.737932865983312,
+      "learning_rate": 1.0752213193472841e-05,
+      "loss": 0.8657,
+      "step": 15143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.255936526345739,
+      "learning_rate": 1.0748786424553275e-05,
+      "loss": 0.7306,
+      "step": 15144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.411574084234449,
+      "learning_rate": 1.074536005223917e-05,
+      "loss": 1.0234,
+      "step": 15145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.429693617413394,
+      "learning_rate": 1.0741934076625895e-05,
+      "loss": 0.9248,
+      "step": 15146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.907019400527617,
+      "learning_rate": 1.0738508497808797e-05,
+      "loss": 1.0957,
+      "step": 15147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.257019959841094,
+      "learning_rate": 1.0735083315883201e-05,
+      "loss": 1.1855,
+      "step": 15148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3062133910212075,
+      "learning_rate": 1.073165853094443e-05,
+      "loss": 0.8687,
+      "step": 15149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.657337115180871,
+      "learning_rate": 1.0728234143087785e-05,
+      "loss": 1.6133,
+      "step": 15150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3626613339836844,
+      "learning_rate": 1.0724810152408581e-05,
+      "loss": 1.0371,
+      "step": 15151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.693560663990723,
+      "learning_rate": 1.0721386559002103e-05,
+      "loss": 1.0117,
+      "step": 15152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1358700030698095,
+      "learning_rate": 1.071796336296362e-05,
+      "loss": 1.1055,
+      "step": 15153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.1815724349517,
+      "learning_rate": 1.0714540564388395e-05,
+      "loss": 1.3184,
+      "step": 15154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.832268814120053,
+      "learning_rate": 1.0711118163371689e-05,
+      "loss": 0.4844,
+      "step": 15155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.440041122246339,
+      "learning_rate": 1.0707696160008759e-05,
+      "loss": 0.6943,
+      "step": 15156
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.4872554871826695,
+      "learning_rate": 1.0704274554394825e-05,
+      "loss": 1.0312,
+      "step": 15157
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.011190999024303,
+      "learning_rate": 1.0700853346625104e-05,
+      "loss": 0.7656,
+      "step": 15158
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.3038200906748365,
+      "learning_rate": 1.0697432536794819e-05,
+      "loss": 1.5762,
+      "step": 15159
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9346123657857506,
+      "learning_rate": 1.0694012124999165e-05,
+      "loss": 0.8691,
+      "step": 15160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7205527456421956,
+      "learning_rate": 1.0690592111333328e-05,
+      "loss": 0.6484,
+      "step": 15161
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.410366846711072,
+      "learning_rate": 1.0687172495892481e-05,
+      "loss": 1.2646,
+      "step": 15162
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.805264129442627,
+      "learning_rate": 1.0683753278771801e-05,
+      "loss": 1.2656,
+      "step": 15163
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.849735234455695,
+      "learning_rate": 1.0680334460066443e-05,
+      "loss": 1.479,
+      "step": 15164
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.616982769373234,
+      "learning_rate": 1.067691603987154e-05,
+      "loss": 1.2461,
+      "step": 15165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9150745050544,
+      "learning_rate": 1.0673498018282239e-05,
+      "loss": 0.8682,
+      "step": 15166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.475690698820596,
+      "learning_rate": 1.0670080395393647e-05,
+      "loss": 1.6172,
+      "step": 15167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6160999088410453,
+      "learning_rate": 1.0666663171300895e-05,
+      "loss": 1.3623,
+      "step": 15168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.455360900083367,
+      "learning_rate": 1.0663246346099067e-05,
+      "loss": 1.2676,
+      "step": 15169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9399023236923345,
+      "learning_rate": 1.065982991988326e-05,
+      "loss": 1.334,
+      "step": 15170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6940188634391857,
+      "learning_rate": 1.0656413892748541e-05,
+      "loss": 0.8828,
+      "step": 15171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.3243769106529575,
+      "learning_rate": 1.065299826478999e-05,
+      "loss": 1.2539,
+      "step": 15172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1068646221872838,
+      "learning_rate": 1.0649583036102659e-05,
+      "loss": 1.2344,
+      "step": 15173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.635863124932716,
+      "learning_rate": 1.0646168206781578e-05,
+      "loss": 1.3496,
+      "step": 15174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.55705939725261,
+      "learning_rate": 1.0642753776921805e-05,
+      "loss": 1.1914,
+      "step": 15175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.166537750698564,
+      "learning_rate": 1.0639339746618338e-05,
+      "loss": 1.0508,
+      "step": 15176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.418567807860603,
+      "learning_rate": 1.063592611596621e-05,
+      "loss": 2.0439,
+      "step": 15177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.156558802509883,
+      "learning_rate": 1.0632512885060412e-05,
+      "loss": 1.5332,
+      "step": 15178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.644511505909724,
+      "learning_rate": 1.0629100053995919e-05,
+      "loss": 1.2129,
+      "step": 15179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.118910647803258,
+      "learning_rate": 1.0625687622867731e-05,
+      "loss": 1.3281,
+      "step": 15180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.207713972284546,
+      "learning_rate": 1.0622275591770805e-05,
+      "loss": 1.6113,
+      "step": 15181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4875982524575382,
+      "learning_rate": 1.0618863960800097e-05,
+      "loss": 0.8787,
+      "step": 15182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4729423962886936,
+      "learning_rate": 1.061545273005054e-05,
+      "loss": 1.1094,
+      "step": 15183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9014454166287997,
+      "learning_rate": 1.0612041899617078e-05,
+      "loss": 0.3827,
+      "step": 15184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.352626116429739,
+      "learning_rate": 1.0608631469594643e-05,
+      "loss": 0.9033,
+      "step": 15185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.3557866041820725,
+      "learning_rate": 1.0605221440078133e-05,
+      "loss": 1.5098,
+      "step": 15186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.515256203101869,
+      "learning_rate": 1.0601811811162455e-05,
+      "loss": 1.0439,
+      "step": 15187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.735736268299176,
+      "learning_rate": 1.059840258294248e-05,
+      "loss": 0.8014,
+      "step": 15188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.3655228566760105,
+      "learning_rate": 1.0594993755513106e-05,
+      "loss": 1.293,
+      "step": 15189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.289091953293573,
+      "learning_rate": 1.0591585328969192e-05,
+      "loss": 1.0342,
+      "step": 15190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6567979135325346,
+      "learning_rate": 1.0588177303405594e-05,
+      "loss": 1.0029,
+      "step": 15191
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.4140616357853135,
+      "learning_rate": 1.0584769678917145e-05,
+      "loss": 1.1777,
+      "step": 15192
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.98284521389702,
+      "learning_rate": 1.058136245559869e-05,
+      "loss": 1.209,
+      "step": 15193
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.926128618804596,
+      "learning_rate": 1.0577955633545054e-05,
+      "loss": 0.8672,
+      "step": 15194
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5350998679032823,
+      "learning_rate": 1.0574549212851042e-05,
+      "loss": 1.1777,
+      "step": 15195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.557486637221738,
+      "learning_rate": 1.0571143193611444e-05,
+      "loss": 1.1875,
+      "step": 15196
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8076873292868787,
+      "learning_rate": 1.0567737575921065e-05,
+      "loss": 1.3672,
+      "step": 15197
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296139034773482,
+      "learning_rate": 1.0564332359874676e-05,
+      "loss": 1.0254,
+      "step": 15198
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5428396788179857,
+      "learning_rate": 1.056092754556704e-05,
+      "loss": 1.4492,
+      "step": 15199
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.048595095103632,
+      "learning_rate": 1.0557523133092903e-05,
+      "loss": 1.3828,
+      "step": 15200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.11183689702848,
+      "learning_rate": 1.0554119122547026e-05,
+      "loss": 1.2949,
+      "step": 15201
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.611793356790353,
+      "learning_rate": 1.0550715514024122e-05,
+      "loss": 0.4932,
+      "step": 15202
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1314418772006043,
+      "learning_rate": 1.0547312307618937e-05,
+      "loss": 1.457,
+      "step": 15203
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.337850954361366,
+      "learning_rate": 1.0543909503426164e-05,
+      "loss": 1.0459,
+      "step": 15204
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.77487176053869,
+      "learning_rate": 1.0540507101540495e-05,
+      "loss": 0.6963,
+      "step": 15205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7803921557928786,
+      "learning_rate": 1.0537105102056635e-05,
+      "loss": 0.7003,
+      "step": 15206
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1201772478080887,
+      "learning_rate": 1.0533703505069257e-05,
+      "loss": 0.9402,
+      "step": 15207
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9936944385786795,
+      "learning_rate": 1.0530302310673015e-05,
+      "loss": 0.7397,
+      "step": 15208
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.456581888773913,
+      "learning_rate": 1.0526901518962565e-05,
+      "loss": 0.8984,
+      "step": 15209
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8675820799615948,
+      "learning_rate": 1.0523501130032559e-05,
+      "loss": 0.4398,
+      "step": 15210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.341545135229145,
+      "learning_rate": 1.0520101143977614e-05,
+      "loss": 1.3999,
+      "step": 15211
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3249151645072477,
+      "learning_rate": 1.0516701560892373e-05,
+      "loss": 1.5762,
+      "step": 15212
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.059340436757447,
+      "learning_rate": 1.051330238087143e-05,
+      "loss": 1.3389,
+      "step": 15213
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.623715040783655,
+      "learning_rate": 1.0509903604009371e-05,
+      "loss": 1.0886,
+      "step": 15214
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.234738267469399,
+      "learning_rate": 1.0506505230400809e-05,
+      "loss": 1.2461,
+      "step": 15215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2678558941763667,
+      "learning_rate": 1.0503107260140305e-05,
+      "loss": 0.5005,
+      "step": 15216
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8734236002799007,
+      "learning_rate": 1.0499709693322412e-05,
+      "loss": 1.2656,
+      "step": 15217
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.325791378064065,
+      "learning_rate": 1.0496312530041708e-05,
+      "loss": 1.2061,
+      "step": 15218
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9116428341806366,
+      "learning_rate": 1.0492915770392722e-05,
+      "loss": 0.8438,
+      "step": 15219
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.629472438019385,
+      "learning_rate": 1.0489519414469976e-05,
+      "loss": 0.9966,
+      "step": 15220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.359263652879921,
+      "learning_rate": 1.0486123462368002e-05,
+      "loss": 0.9502,
+      "step": 15221
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2152507241641795,
+      "learning_rate": 1.0482727914181297e-05,
+      "loss": 0.7214,
+      "step": 15222
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.787124772754418,
+      "learning_rate": 1.0479332770004375e-05,
+      "loss": 1.3008,
+      "step": 15223
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.146565568549395,
+      "learning_rate": 1.047593802993171e-05,
+      "loss": 0.8428,
+      "step": 15224
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.032270432950858,
+      "learning_rate": 1.0472543694057773e-05,
+      "loss": 1.7832,
+      "step": 15225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3290374811022057,
+      "learning_rate": 1.0469149762477026e-05,
+      "loss": 1.1572,
+      "step": 15226
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1666771692852858,
+      "learning_rate": 1.0465756235283933e-05,
+      "loss": 1.1152,
+      "step": 15227
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.595603796453633,
+      "learning_rate": 1.0462363112572929e-05,
+      "loss": 0.8066,
+      "step": 15228
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3735862338768436,
+      "learning_rate": 1.0458970394438431e-05,
+      "loss": 0.5137,
+      "step": 15229
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.735055021468304,
+      "learning_rate": 1.0455578080974873e-05,
+      "loss": 1.0459,
+      "step": 15230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9656756092251277,
+      "learning_rate": 1.0452186172276652e-05,
+      "loss": 0.8965,
+      "step": 15231
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6001614640372273,
+      "learning_rate": 1.0448794668438176e-05,
+      "loss": 0.8457,
+      "step": 15232
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8461243495086808,
+      "learning_rate": 1.0445403569553815e-05,
+      "loss": 0.667,
+      "step": 15233
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0904480817944804,
+      "learning_rate": 1.0442012875717952e-05,
+      "loss": 0.854,
+      "step": 15234
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3448577127106445,
+      "learning_rate": 1.0438622587024934e-05,
+      "loss": 1.4355,
+      "step": 15235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.950512985928892,
+      "learning_rate": 1.0435232703569131e-05,
+      "loss": 0.8491,
+      "step": 15236
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5956256932405335,
+      "learning_rate": 1.0431843225444873e-05,
+      "loss": 0.8271,
+      "step": 15237
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3840028285201855,
+      "learning_rate": 1.0428454152746477e-05,
+      "loss": 0.8472,
+      "step": 15238
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.882797447930301,
+      "learning_rate": 1.0425065485568271e-05,
+      "loss": 1.1484,
+      "step": 15239
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.195671826143942,
+      "learning_rate": 1.0421677224004569e-05,
+      "loss": 1.2949,
+      "step": 15240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.785937117739743,
+      "learning_rate": 1.0418289368149653e-05,
+      "loss": 1.4795,
+      "step": 15241
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.297840023444289,
+      "learning_rate": 1.041490191809781e-05,
+      "loss": 1.6646,
+      "step": 15242
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.486374056767162,
+      "learning_rate": 1.0411514873943298e-05,
+      "loss": 0.8271,
+      "step": 15243
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.162807830774641,
+      "learning_rate": 1.0408128235780399e-05,
+      "loss": 1.0205,
+      "step": 15244
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8182140117790797,
+      "learning_rate": 1.0404742003703349e-05,
+      "loss": 0.8037,
+      "step": 15245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.4255983023611964,
+      "learning_rate": 1.0401356177806388e-05,
+      "loss": 0.7715,
+      "step": 15246
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.120714377023541,
+      "learning_rate": 1.0397970758183732e-05,
+      "loss": 1.2852,
+      "step": 15247
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.59333145257576,
+      "learning_rate": 1.0394585744929605e-05,
+      "loss": 0.9971,
+      "step": 15248
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.22248150451997,
+      "learning_rate": 1.039120113813822e-05,
+      "loss": 1.2773,
+      "step": 15249
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.034513945120288,
+      "learning_rate": 1.0387816937903761e-05,
+      "loss": 1.2656,
+      "step": 15250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.405013026561076,
+      "learning_rate": 1.0384433144320408e-05,
+      "loss": 1.2676,
+      "step": 15251
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.548499453771,
+      "learning_rate": 1.038104975748232e-05,
+      "loss": 0.9307,
+      "step": 15252
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3010630928816087,
+      "learning_rate": 1.0377666777483674e-05,
+      "loss": 0.79,
+      "step": 15253
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.47661311080612,
+      "learning_rate": 1.0374284204418613e-05,
+      "loss": 1.8477,
+      "step": 15254
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5003601978785808,
+      "learning_rate": 1.0370902038381267e-05,
+      "loss": 1.0117,
+      "step": 15255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.77177899698659,
+      "learning_rate": 1.0367520279465751e-05,
+      "loss": 0.8877,
+      "step": 15256
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.022295065966336,
+      "learning_rate": 1.036413892776619e-05,
+      "loss": 0.7207,
+      "step": 15257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.167636575208971,
+      "learning_rate": 1.0360757983376693e-05,
+      "loss": 0.8955,
+      "step": 15258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8331601333038665,
+      "learning_rate": 1.0357377446391343e-05,
+      "loss": 1.373,
+      "step": 15259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3187661553113585,
+      "learning_rate": 1.0353997316904204e-05,
+      "loss": 0.8589,
+      "step": 15260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.023395545518516,
+      "learning_rate": 1.0350617595009372e-05,
+      "loss": 0.7764,
+      "step": 15261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.015180643974072,
+      "learning_rate": 1.0347238280800885e-05,
+      "loss": 1.2402,
+      "step": 15262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.296197804380546,
+      "learning_rate": 1.0343859374372795e-05,
+      "loss": 1.8457,
+      "step": 15263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1949294030832722,
+      "learning_rate": 1.034048087581912e-05,
+      "loss": 1.2666,
+      "step": 15264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.464011444560139,
+      "learning_rate": 1.0337102785233905e-05,
+      "loss": 0.9561,
+      "step": 15265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.5292418546913735,
+      "learning_rate": 1.0333725102711148e-05,
+      "loss": 1.6426,
+      "step": 15266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2970064086691804,
+      "learning_rate": 1.0330347828344841e-05,
+      "loss": 1.0927,
+      "step": 15267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.116690618365542,
+      "learning_rate": 1.0326970962228994e-05,
+      "loss": 1.3242,
+      "step": 15268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.691424055552722,
+      "learning_rate": 1.0323594504457562e-05,
+      "loss": 1.1104,
+      "step": 15269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3347946460645668,
+      "learning_rate": 1.032021845512453e-05,
+      "loss": 0.8555,
+      "step": 15270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.531931373347153,
+      "learning_rate": 1.031684281432384e-05,
+      "loss": 1.2852,
+      "step": 15271
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1610340728474338,
+      "learning_rate": 1.031346758214944e-05,
+      "loss": 1.4219,
+      "step": 15272
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.133149925572621,
+      "learning_rate": 1.0310092758695247e-05,
+      "loss": 0.8687,
+      "step": 15273
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.590918578254976,
+      "learning_rate": 1.0306718344055202e-05,
+      "loss": 1.4863,
+      "step": 15274
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.249920617366272,
+      "learning_rate": 1.0303344338323203e-05,
+      "loss": 0.7681,
+      "step": 15275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4744396114637675,
+      "learning_rate": 1.029997074159314e-05,
+      "loss": 1.0762,
+      "step": 15276
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1880941401563425,
+      "learning_rate": 1.0296597553958906e-05,
+      "loss": 0.8009,
+      "step": 15277
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1685208028935428,
+      "learning_rate": 1.0293224775514385e-05,
+      "loss": 1.1416,
+      "step": 15278
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2876090782258327,
+      "learning_rate": 1.0289852406353435e-05,
+      "loss": 1.0322,
+      "step": 15279
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.563960662684645,
+      "learning_rate": 1.0286480446569899e-05,
+      "loss": 0.7031,
+      "step": 15280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.721322072322085,
+      "learning_rate": 1.0283108896257615e-05,
+      "loss": 1.5986,
+      "step": 15281
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.046770925456095,
+      "learning_rate": 1.027973775551043e-05,
+      "loss": 1.105,
+      "step": 15282
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9273842822894185,
+      "learning_rate": 1.0276367024422149e-05,
+      "loss": 0.8545,
+      "step": 15283
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2337110906882782,
+      "learning_rate": 1.027299670308658e-05,
+      "loss": 1.3193,
+      "step": 15284
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.989268021564566,
+      "learning_rate": 1.0269626791597509e-05,
+      "loss": 1.2539,
+      "step": 15285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.016554625475252,
+      "learning_rate": 1.0266257290048726e-05,
+      "loss": 1.0801,
+      "step": 15286
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.902277888542425,
+      "learning_rate": 1.0262888198534015e-05,
+      "loss": 0.8486,
+      "step": 15287
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.164345336240473,
+      "learning_rate": 1.0259519517147126e-05,
+      "loss": 0.6816,
+      "step": 15288
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7281759565385815,
+      "learning_rate": 1.0256151245981806e-05,
+      "loss": 1.498,
+      "step": 15289
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.331175783014021,
+      "learning_rate": 1.025278338513179e-05,
+      "loss": 0.6951,
+      "step": 15290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.403827278737067,
+      "learning_rate": 1.0249415934690811e-05,
+      "loss": 1.6055,
+      "step": 15291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7840173205830656,
+      "learning_rate": 1.0246048894752589e-05,
+      "loss": 0.3066,
+      "step": 15292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9939511578304576,
+      "learning_rate": 1.0242682265410817e-05,
+      "loss": 0.959,
+      "step": 15293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.060466081146668,
+      "learning_rate": 1.023931604675917e-05,
+      "loss": 0.7568,
+      "step": 15294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4274969989981994,
+      "learning_rate": 1.0235950238891367e-05,
+      "loss": 0.8906,
+      "step": 15295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9277955103931435,
+      "learning_rate": 1.0232584841901058e-05,
+      "loss": 1.9746,
+      "step": 15296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.738175459490196,
+      "learning_rate": 1.0229219855881903e-05,
+      "loss": 1.1289,
+      "step": 15297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6361019967335517,
+      "learning_rate": 1.0225855280927537e-05,
+      "loss": 1.1777,
+      "step": 15298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5407981222366773,
+      "learning_rate": 1.022249111713161e-05,
+      "loss": 0.501,
+      "step": 15299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9543339913727706,
+      "learning_rate": 1.0219127364587743e-05,
+      "loss": 0.7939,
+      "step": 15300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8570403500447523,
+      "learning_rate": 1.0215764023389544e-05,
+      "loss": 0.7412,
+      "step": 15301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.311764185531459,
+      "learning_rate": 1.0212401093630602e-05,
+      "loss": 0.6216,
+      "step": 15302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6672513254205856,
+      "learning_rate": 1.0209038575404528e-05,
+      "loss": 0.6343,
+      "step": 15303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.151534854486913,
+      "learning_rate": 1.0205676468804881e-05,
+      "loss": 1.04,
+      "step": 15304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1590882183435234,
+      "learning_rate": 1.0202314773925243e-05,
+      "loss": 1.2622,
+      "step": 15305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.285979010238282,
+      "learning_rate": 1.0198953490859159e-05,
+      "loss": 1.4385,
+      "step": 15306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.756140905424886,
+      "learning_rate": 1.0195592619700167e-05,
+      "loss": 1.5391,
+      "step": 15307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3904049731433414,
+      "learning_rate": 1.0192232160541815e-05,
+      "loss": 0.6754,
+      "step": 15308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.133266806065312,
+      "learning_rate": 1.0188872113477612e-05,
+      "loss": 0.9111,
+      "step": 15309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.063383433401874,
+      "learning_rate": 1.018551247860107e-05,
+      "loss": 1.0576,
+      "step": 15310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.94553639499427,
+      "learning_rate": 1.0182153256005675e-05,
+      "loss": 1.3242,
+      "step": 15311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.725589174045239,
+      "learning_rate": 1.0178794445784931e-05,
+      "loss": 1.5957,
+      "step": 15312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1487600743277846,
+      "learning_rate": 1.0175436048032293e-05,
+      "loss": 0.915,
+      "step": 15313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.278048597523693,
+      "learning_rate": 1.0172078062841248e-05,
+      "loss": 0.542,
+      "step": 15314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8730115233527815,
+      "learning_rate": 1.0168720490305223e-05,
+      "loss": 0.521,
+      "step": 15315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6052202967008453,
+      "learning_rate": 1.0165363330517677e-05,
+      "loss": 1.1206,
+      "step": 15316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2202893142509716,
+      "learning_rate": 1.0162006583572031e-05,
+      "loss": 0.9004,
+      "step": 15317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8190191037064585,
+      "learning_rate": 1.0158650249561702e-05,
+      "loss": 1.0103,
+      "step": 15318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9856930658082887,
+      "learning_rate": 1.0155294328580087e-05,
+      "loss": 0.6072,
+      "step": 15319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5888020827428084,
+      "learning_rate": 1.0151938820720595e-05,
+      "loss": 0.3169,
+      "step": 15320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.254826890088952,
+      "learning_rate": 1.01485837260766e-05,
+      "loss": 1.2168,
+      "step": 15321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4936443626686904,
+      "learning_rate": 1.0145229044741464e-05,
+      "loss": 1.1189,
+      "step": 15322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.946233569635069,
+      "learning_rate": 1.0141874776808569e-05,
+      "loss": 1.3125,
+      "step": 15323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.147967813056101,
+      "learning_rate": 1.0138520922371238e-05,
+      "loss": 1.3164,
+      "step": 15324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7762301834227956,
+      "learning_rate": 1.013516748152283e-05,
+      "loss": 0.619,
+      "step": 15325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.422413911887946,
+      "learning_rate": 1.0131814454356659e-05,
+      "loss": 1.4805,
+      "step": 15326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6149036064397633,
+      "learning_rate": 1.0128461840966038e-05,
+      "loss": 1.3145,
+      "step": 15327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3610789228252345,
+      "learning_rate": 1.0125109641444263e-05,
+      "loss": 1.0723,
+      "step": 15328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.9409563545811634,
+      "learning_rate": 1.0121757855884639e-05,
+      "loss": 0.708,
+      "step": 15329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.033796035254961,
+      "learning_rate": 1.0118406484380439e-05,
+      "loss": 1.498,
+      "step": 15330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.937392699534207,
+      "learning_rate": 1.0115055527024917e-05,
+      "loss": 0.8396,
+      "step": 15331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.108717805813065,
+      "learning_rate": 1.011170498391135e-05,
+      "loss": 1.3516,
+      "step": 15332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.660729882769711,
+      "learning_rate": 1.010835485513296e-05,
+      "loss": 1.2949,
+      "step": 15333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.415546107144213,
+      "learning_rate": 1.0105005140783006e-05,
+      "loss": 1.1914,
+      "step": 15334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6425950118995294,
+      "learning_rate": 1.010165584095469e-05,
+      "loss": 1.3887,
+      "step": 15335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3902195605376724,
+      "learning_rate": 1.0098306955741221e-05,
+      "loss": 0.8867,
+      "step": 15336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1359419947631433,
+      "learning_rate": 1.0094958485235811e-05,
+      "loss": 0.8867,
+      "step": 15337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.893175784639184,
+      "learning_rate": 1.009161042953164e-05,
+      "loss": 1.416,
+      "step": 15338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.1020172738948855,
+      "learning_rate": 1.008826278872188e-05,
+      "loss": 0.6836,
+      "step": 15339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.910751302418544,
+      "learning_rate": 1.0084915562899689e-05,
+      "loss": 1.3945,
+      "step": 15340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.043290955608931,
+      "learning_rate": 1.0081568752158224e-05,
+      "loss": 1.4082,
+      "step": 15341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7284181941992953,
+      "learning_rate": 1.0078222356590637e-05,
+      "loss": 0.8906,
+      "step": 15342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4626462194919745,
+      "learning_rate": 1.0074876376290047e-05,
+      "loss": 0.8174,
+      "step": 15343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.560208389175029,
+      "learning_rate": 1.0071530811349569e-05,
+      "loss": 1.5352,
+      "step": 15344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.208508454582781,
+      "learning_rate": 1.0068185661862306e-05,
+      "loss": 0.7646,
+      "step": 15345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3337593257413283,
+      "learning_rate": 1.0064840927921365e-05,
+      "loss": 1.7324,
+      "step": 15346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4559470680298223,
+      "learning_rate": 1.006149660961982e-05,
+      "loss": 1.2539,
+      "step": 15347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.243595268072378,
+      "learning_rate": 1.0058152707050744e-05,
+      "loss": 0.5688,
+      "step": 15348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.499874588681793,
+      "learning_rate": 1.0054809220307185e-05,
+      "loss": 0.7081,
+      "step": 15349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9966222731485017,
+      "learning_rate": 1.0051466149482203e-05,
+      "loss": 1.373,
+      "step": 15350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.852397864166281,
+      "learning_rate": 1.004812349466884e-05,
+      "loss": 1.1787,
+      "step": 15351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.539649965842759,
+      "learning_rate": 1.0044781255960115e-05,
+      "loss": 1.0254,
+      "step": 15352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9030520546214325,
+      "learning_rate": 1.004143943344903e-05,
+      "loss": 0.6486,
+      "step": 15353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.339467828831474,
+      "learning_rate": 1.0038098027228603e-05,
+      "loss": 0.6035,
+      "step": 15354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5514315233410323,
+      "learning_rate": 1.0034757037391817e-05,
+      "loss": 0.6953,
+      "step": 15355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.396180125030119,
+      "learning_rate": 1.0031416464031654e-05,
+      "loss": 1.084,
+      "step": 15356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.102777066294344,
+      "learning_rate": 1.0028076307241063e-05,
+      "loss": 0.8149,
+      "step": 15357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3833669548953247,
+      "learning_rate": 1.0024736567113026e-05,
+      "loss": 0.7366,
+      "step": 15358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.040133009169706,
+      "learning_rate": 1.0021397243740463e-05,
+      "loss": 0.9668,
+      "step": 15359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4773512104827944,
+      "learning_rate": 1.0018058337216327e-05,
+      "loss": 1.0923,
+      "step": 15360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0244605726263143,
+      "learning_rate": 1.0014719847633527e-05,
+      "loss": 0.5947,
+      "step": 15361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5526164114445065,
+      "learning_rate": 1.0011381775084966e-05,
+      "loss": 0.9736,
+      "step": 15362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0152718925291735,
+      "learning_rate": 1.0008044119663557e-05,
+      "loss": 0.7461,
+      "step": 15363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.020289948799104,
+      "learning_rate": 1.0004706881462176e-05,
+      "loss": 1.1523,
+      "step": 15364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.262765592719227,
+      "learning_rate": 1.0001370060573698e-05,
+      "loss": 1.0186,
+      "step": 15365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4005202405060753,
+      "learning_rate": 9.998033657090978e-06,
+      "loss": 1.0566,
+      "step": 15366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.41075572744827,
+      "learning_rate": 9.994697671106882e-06,
+      "loss": 0.6931,
+      "step": 15367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.925935096923076,
+      "learning_rate": 9.991362102714239e-06,
+      "loss": 1.0361,
+      "step": 15368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3013809536500562,
+      "learning_rate": 9.988026952005872e-06,
+      "loss": 1.4062,
+      "step": 15369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.014165169926201,
+      "learning_rate": 9.98469221907461e-06,
+      "loss": 1.1045,
+      "step": 15370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.814125354325421,
+      "learning_rate": 9.981357904013245e-06,
+      "loss": 1.3438,
+      "step": 15371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4564700948160136,
+      "learning_rate": 9.978024006914583e-06,
+      "loss": 0.6079,
+      "step": 15372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.692537911468039,
+      "learning_rate": 9.974690527871397e-06,
+      "loss": 0.8438,
+      "step": 15373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5178332380804562,
+      "learning_rate": 9.97135746697645e-06,
+      "loss": 0.835,
+      "step": 15374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.569068491698706,
+      "learning_rate": 9.968024824322514e-06,
+      "loss": 1.1055,
+      "step": 15375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.138231272740118,
+      "learning_rate": 9.964692600002327e-06,
+      "loss": 2.1992,
+      "step": 15376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.394480768514826,
+      "learning_rate": 9.961360794108624e-06,
+      "loss": 0.7007,
+      "step": 15377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.233992562478958,
+      "learning_rate": 9.958029406734118e-06,
+      "loss": 1.146,
+      "step": 15378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0690015949979723,
+      "learning_rate": 9.954698437971532e-06,
+      "loss": 1.2695,
+      "step": 15379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.233158933386603,
+      "learning_rate": 9.951367887913574e-06,
+      "loss": 1.0273,
+      "step": 15380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2385886935630186,
+      "learning_rate": 9.948037756652917e-06,
+      "loss": 1.2012,
+      "step": 15381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.130222410670045,
+      "learning_rate": 9.944708044282244e-06,
+      "loss": 0.9482,
+      "step": 15382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.596340035285614,
+      "learning_rate": 9.941378750894209e-06,
+      "loss": 0.7969,
+      "step": 15383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1743949706324672,
+      "learning_rate": 9.938049876581481e-06,
+      "loss": 1.1816,
+      "step": 15384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0679049383006016,
+      "learning_rate": 9.934721421436693e-06,
+      "loss": 0.6914,
+      "step": 15385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1591115567255867,
+      "learning_rate": 9.931393385552476e-06,
+      "loss": 0.9658,
+      "step": 15386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6384333615802498,
+      "learning_rate": 9.928065769021436e-06,
+      "loss": 0.9766,
+      "step": 15387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0112997563136417,
+      "learning_rate": 9.92473857193619e-06,
+      "loss": 1.0732,
+      "step": 15388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.350335393468769,
+      "learning_rate": 9.921411794389343e-06,
+      "loss": 0.9082,
+      "step": 15389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.411289041617392,
+      "learning_rate": 9.918085436473465e-06,
+      "loss": 0.4722,
+      "step": 15390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9215970962693052,
+      "learning_rate": 9.91475949828113e-06,
+      "loss": 0.8018,
+      "step": 15391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.077088002223091,
+      "learning_rate": 9.911433979904891e-06,
+      "loss": 1.1943,
+      "step": 15392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9627110144554734,
+      "learning_rate": 9.908108881437309e-06,
+      "loss": 0.8301,
+      "step": 15393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9259195336143395,
+      "learning_rate": 9.90478420297091e-06,
+      "loss": 1.3735,
+      "step": 15394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.805777672624514,
+      "learning_rate": 9.901459944598215e-06,
+      "loss": 0.9233,
+      "step": 15395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.139977687343361,
+      "learning_rate": 9.898136106411742e-06,
+      "loss": 0.6885,
+      "step": 15396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.688545821988603,
+      "learning_rate": 9.894812688504004e-06,
+      "loss": 1.5039,
+      "step": 15397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2265791996484454,
+      "learning_rate": 9.891489690967479e-06,
+      "loss": 0.7253,
+      "step": 15398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3054879303612155,
+      "learning_rate": 9.888167113894645e-06,
+      "loss": 0.9868,
+      "step": 15399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.208555827751619,
+      "learning_rate": 9.884844957377958e-06,
+      "loss": 1.4492,
+      "step": 15400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9482240730024825,
+      "learning_rate": 9.881523221509891e-06,
+      "loss": 1.6328,
+      "step": 15401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.710429395724864,
+      "learning_rate": 9.878201906382881e-06,
+      "loss": 0.6045,
+      "step": 15402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5484218463615753,
+      "learning_rate": 9.874881012089355e-06,
+      "loss": 0.9727,
+      "step": 15403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5319938037790473,
+      "learning_rate": 9.871560538721722e-06,
+      "loss": 1.3359,
+      "step": 15404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.918351719786454,
+      "learning_rate": 9.868240486372413e-06,
+      "loss": 1.166,
+      "step": 15405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.239702637531177,
+      "learning_rate": 9.8649208551338e-06,
+      "loss": 1.0879,
+      "step": 15406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.029437226787882,
+      "learning_rate": 9.861601645098283e-06,
+      "loss": 0.9092,
+      "step": 15407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.422972202949231,
+      "learning_rate": 9.858282856358237e-06,
+      "loss": 0.9795,
+      "step": 15408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.746221544827203,
+      "learning_rate": 9.854964489006e-06,
+      "loss": 0.8428,
+      "step": 15409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.031063858285102,
+      "learning_rate": 9.851646543133946e-06,
+      "loss": 1.2451,
+      "step": 15410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.19938225147866,
+      "learning_rate": 9.848329018834404e-06,
+      "loss": 1.7656,
+      "step": 15411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.927691520052242,
+      "learning_rate": 9.845011916199696e-06,
+      "loss": 1.0654,
+      "step": 15412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.83102641669007,
+      "learning_rate": 9.84169523532213e-06,
+      "loss": 1.7012,
+      "step": 15413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.758346765820772,
+      "learning_rate": 9.83837897629402e-06,
+      "loss": 2.1055,
+      "step": 15414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.739218846244486,
+      "learning_rate": 9.835063139207643e-06,
+      "loss": 1.4121,
+      "step": 15415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.735736093999782,
+      "learning_rate": 9.831747724155293e-06,
+      "loss": 1.2285,
+      "step": 15416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.548974250618274,
+      "learning_rate": 9.828432731229223e-06,
+      "loss": 1.5957,
+      "step": 15417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6906760322790912,
+      "learning_rate": 9.825118160521698e-06,
+      "loss": 0.2122,
+      "step": 15418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.600543028883871,
+      "learning_rate": 9.82180401212496e-06,
+      "loss": 0.9336,
+      "step": 15419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.660451708158936,
+      "learning_rate": 9.818490286131238e-06,
+      "loss": 1.0569,
+      "step": 15420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.308106659311028,
+      "learning_rate": 9.815176982632742e-06,
+      "loss": 1.3105,
+      "step": 15421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.436334869487969,
+      "learning_rate": 9.811864101721696e-06,
+      "loss": 0.9268,
+      "step": 15422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.764501397289655,
+      "learning_rate": 9.80855164349029e-06,
+      "loss": 0.6816,
+      "step": 15423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.316372490004829,
+      "learning_rate": 9.805239608030698e-06,
+      "loss": 1.0674,
+      "step": 15424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.493844489108498,
+      "learning_rate": 9.801927995435112e-06,
+      "loss": 0.5134,
+      "step": 15425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3463401215387303,
+      "learning_rate": 9.798616805795672e-06,
+      "loss": 1.3086,
+      "step": 15426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.275004228196072,
+      "learning_rate": 9.79530603920455e-06,
+      "loss": 1.5762,
+      "step": 15427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.172136117494424,
+      "learning_rate": 9.791995695753866e-06,
+      "loss": 1.2773,
+      "step": 15428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.838708628992767,
+      "learning_rate": 9.788685775535756e-06,
+      "loss": 1.3926,
+      "step": 15429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.343396269930742,
+      "learning_rate": 9.785376278642314e-06,
+      "loss": 0.9834,
+      "step": 15430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.783289655071749,
+      "learning_rate": 9.782067205165668e-06,
+      "loss": 1.0029,
+      "step": 15431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.46940730285175,
+      "learning_rate": 9.778758555197898e-06,
+      "loss": 0.9639,
+      "step": 15432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0335702408945946,
+      "learning_rate": 9.77545032883107e-06,
+      "loss": 0.6934,
+      "step": 15433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8853136446115126,
+      "learning_rate": 9.77214252615726e-06,
+      "loss": 0.9902,
+      "step": 15434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6176354779266524,
+      "learning_rate": 9.768835147268534e-06,
+      "loss": 0.2781,
+      "step": 15435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.447728675919941,
+      "learning_rate": 9.76552819225693e-06,
+      "loss": 1.4883,
+      "step": 15436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7179394971781554,
+      "learning_rate": 9.762221661214468e-06,
+      "loss": 1.2676,
+      "step": 15437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6058529594844484,
+      "learning_rate": 9.75891555423317e-06,
+      "loss": 1.4004,
+      "step": 15438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.075224264503921,
+      "learning_rate": 9.755609871405053e-06,
+      "loss": 0.9072,
+      "step": 15439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.274990010501602,
+      "learning_rate": 9.752304612822107e-06,
+      "loss": 0.9468,
+      "step": 15440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.043681294772907,
+      "learning_rate": 9.748999778576317e-06,
+      "loss": 0.8198,
+      "step": 15441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9929566353493233,
+      "learning_rate": 9.745695368759646e-06,
+      "loss": 0.9912,
+      "step": 15442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9222007202724023,
+      "learning_rate": 9.742391383464064e-06,
+      "loss": 1.6133,
+      "step": 15443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0545160972332157,
+      "learning_rate": 9.739087822781526e-06,
+      "loss": 1.1211,
+      "step": 15444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.809420504487964,
+      "learning_rate": 9.735784686803962e-06,
+      "loss": 1.0098,
+      "step": 15445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.175589250566178,
+      "learning_rate": 9.732481975623294e-06,
+      "loss": 1.5112,
+      "step": 15446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.92470384332445,
+      "learning_rate": 9.72917968933143e-06,
+      "loss": 1.0342,
+      "step": 15447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.082878012174173,
+      "learning_rate": 9.725877828020288e-06,
+      "loss": 0.9116,
+      "step": 15448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.134017750546762,
+      "learning_rate": 9.722576391781746e-06,
+      "loss": 0.8081,
+      "step": 15449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.0349644108870955,
+      "learning_rate": 9.719275380707684e-06,
+      "loss": 1.0615,
+      "step": 15450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.18814285155767,
+      "learning_rate": 9.715974794889962e-06,
+      "loss": 1.1738,
+      "step": 15451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.630384455428958,
+      "learning_rate": 9.712674634420435e-06,
+      "loss": 0.7383,
+      "step": 15452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0387630315816727,
+      "learning_rate": 9.70937489939096e-06,
+      "loss": 1.0195,
+      "step": 15453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1344890781819266,
+      "learning_rate": 9.706075589893357e-06,
+      "loss": 1.2578,
+      "step": 15454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8530740924822875,
+      "learning_rate": 9.702776706019435e-06,
+      "loss": 0.8574,
+      "step": 15455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.81336758265808,
+      "learning_rate": 9.699478247861018e-06,
+      "loss": 0.6821,
+      "step": 15456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7340518433483845,
+      "learning_rate": 9.696180215509892e-06,
+      "loss": 0.2754,
+      "step": 15457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.093908389349224,
+      "learning_rate": 9.692882609057844e-06,
+      "loss": 1.001,
+      "step": 15458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0042367406108865,
+      "learning_rate": 9.689585428596628e-06,
+      "loss": 1.0132,
+      "step": 15459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.370451794231926,
+      "learning_rate": 9.686288674218027e-06,
+      "loss": 1.3193,
+      "step": 15460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5590115504538944,
+      "learning_rate": 9.682992346013772e-06,
+      "loss": 1.1152,
+      "step": 15461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.680221427577326,
+      "learning_rate": 9.679696444075612e-06,
+      "loss": 0.9912,
+      "step": 15462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0968769339084843,
+      "learning_rate": 9.67640096849526e-06,
+      "loss": 0.8784,
+      "step": 15463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0452520328031816,
+      "learning_rate": 9.673105919364426e-06,
+      "loss": 0.8809,
+      "step": 15464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.392837198983634,
+      "learning_rate": 9.669811296774819e-06,
+      "loss": 1.2178,
+      "step": 15465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.946642925248728,
+      "learning_rate": 9.666517100818127e-06,
+      "loss": 1.6191,
+      "step": 15466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4007400604845373,
+      "learning_rate": 9.663223331586018e-06,
+      "loss": 0.8203,
+      "step": 15467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5256145522579545,
+      "learning_rate": 9.659929989170154e-06,
+      "loss": 0.7944,
+      "step": 15468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9844509320487076,
+      "learning_rate": 9.656637073662203e-06,
+      "loss": 1.5977,
+      "step": 15469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.874338647333714,
+      "learning_rate": 9.653344585153787e-06,
+      "loss": 0.9121,
+      "step": 15470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.6827192955126575,
+      "learning_rate": 9.65005252373655e-06,
+      "loss": 1.5957,
+      "step": 15471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2353325772593733,
+      "learning_rate": 9.646760889502093e-06,
+      "loss": 1.1328,
+      "step": 15472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.020137055823433,
+      "learning_rate": 9.643469682542041e-06,
+      "loss": 1.4258,
+      "step": 15473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.202481982190133,
+      "learning_rate": 9.640178902947977e-06,
+      "loss": 1.3184,
+      "step": 15474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1931934745227797,
+      "learning_rate": 9.63688855081148e-06,
+      "loss": 1.3926,
+      "step": 15475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1061241753805175,
+      "learning_rate": 9.63359862622411e-06,
+      "loss": 0.728,
+      "step": 15476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3798373841211693,
+      "learning_rate": 9.630309129277446e-06,
+      "loss": 1.0176,
+      "step": 15477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.948326935992103,
+      "learning_rate": 9.627020060063019e-06,
+      "loss": 0.7979,
+      "step": 15478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.744938787948642,
+      "learning_rate": 9.623731418672369e-06,
+      "loss": 0.637,
+      "step": 15479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.740229144130818,
+      "learning_rate": 9.620443205197002e-06,
+      "loss": 1.0422,
+      "step": 15480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.781337735876384,
+      "learning_rate": 9.617155419728441e-06,
+      "loss": 0.6699,
+      "step": 15481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.785472091099786,
+      "learning_rate": 9.613868062358192e-06,
+      "loss": 1.0918,
+      "step": 15482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4647921415353924,
+      "learning_rate": 9.610581133177729e-06,
+      "loss": 1.3164,
+      "step": 15483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9752863809131813,
+      "learning_rate": 9.60729463227853e-06,
+      "loss": 0.8811,
+      "step": 15484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.893711445790522,
+      "learning_rate": 9.604008559752042e-06,
+      "loss": 0.9902,
+      "step": 15485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4829482294124916,
+      "learning_rate": 9.600722915689741e-06,
+      "loss": 1.0635,
+      "step": 15486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.098104702485158,
+      "learning_rate": 9.597437700183052e-06,
+      "loss": 1.1133,
+      "step": 15487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.718933274679731,
+      "learning_rate": 9.594152913323402e-06,
+      "loss": 0.4653,
+      "step": 15488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2852065382420723,
+      "learning_rate": 9.590868555202185e-06,
+      "loss": 1.2031,
+      "step": 15489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.082485312767433,
+      "learning_rate": 9.587584625910842e-06,
+      "loss": 1.1396,
+      "step": 15490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.404137910597095,
+      "learning_rate": 9.584301125540742e-06,
+      "loss": 1.0586,
+      "step": 15491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7264423470950336,
+      "learning_rate": 9.581018054183269e-06,
+      "loss": 1.4023,
+      "step": 15492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.095804146272082,
+      "learning_rate": 9.577735411929775e-06,
+      "loss": 1.3105,
+      "step": 15493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8002367328295112,
+      "learning_rate": 9.574453198871638e-06,
+      "loss": 1.0713,
+      "step": 15494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.835261492905777,
+      "learning_rate": 9.571171415100186e-06,
+      "loss": 1.7793,
+      "step": 15495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.558528148013949,
+      "learning_rate": 9.567890060706754e-06,
+      "loss": 1.0195,
+      "step": 15496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5989378556927223,
+      "learning_rate": 9.56460913578265e-06,
+      "loss": 1.2969,
+      "step": 15497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1216895178288215,
+      "learning_rate": 9.56132864041919e-06,
+      "loss": 1.0996,
+      "step": 15498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.783812210628035,
+      "learning_rate": 9.558048574707679e-06,
+      "loss": 1.0537,
+      "step": 15499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0353037365209743,
+      "learning_rate": 9.554768938739386e-06,
+      "loss": 1.0039,
+      "step": 15500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.295307688646357,
+      "learning_rate": 9.551489732605589e-06,
+      "loss": 1.1484,
+      "step": 15501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.875119721336419,
+      "learning_rate": 9.548210956397535e-06,
+      "loss": 1.0381,
+      "step": 15502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.093342208921146,
+      "learning_rate": 9.544932610206486e-06,
+      "loss": 0.7939,
+      "step": 15503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1423872773152377,
+      "learning_rate": 9.541654694123673e-06,
+      "loss": 0.8821,
+      "step": 15504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.647067853957184,
+      "learning_rate": 9.538377208240315e-06,
+      "loss": 1.4141,
+      "step": 15505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.365083648779255,
+      "learning_rate": 9.535100152647617e-06,
+      "loss": 0.6533,
+      "step": 15506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6580770769211246,
+      "learning_rate": 9.531823527436795e-06,
+      "loss": 0.7412,
+      "step": 15507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2526643660481764,
+      "learning_rate": 9.528547332699018e-06,
+      "loss": 1.3184,
+      "step": 15508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.875203909884436,
+      "learning_rate": 9.52527156852548e-06,
+      "loss": 1.2363,
+      "step": 15509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4505156192195006,
+      "learning_rate": 9.521996235007325e-06,
+      "loss": 0.6956,
+      "step": 15510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.378791292667056,
+      "learning_rate": 9.518721332235722e-06,
+      "loss": 0.6723,
+      "step": 15511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1953807019265708,
+      "learning_rate": 9.515446860301803e-06,
+      "loss": 0.5355,
+      "step": 15512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.550027592827601,
+      "learning_rate": 9.512172819296691e-06,
+      "loss": 1.167,
+      "step": 15513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.401201058469893,
+      "learning_rate": 9.508899209311498e-06,
+      "loss": 1.2207,
+      "step": 15514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4022541932189427,
+      "learning_rate": 9.505626030437342e-06,
+      "loss": 0.9678,
+      "step": 15515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9312681282963218,
+      "learning_rate": 9.502353282765306e-06,
+      "loss": 1.1348,
+      "step": 15516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.5076667325733215,
+      "learning_rate": 9.499080966386461e-06,
+      "loss": 1.0405,
+      "step": 15517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0185244502149886,
+      "learning_rate": 9.49580908139189e-06,
+      "loss": 1.4434,
+      "step": 15518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.547721690716073,
+      "learning_rate": 9.492537627872632e-06,
+      "loss": 1.0918,
+      "step": 15519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6558941771096385,
+      "learning_rate": 9.489266605919746e-06,
+      "loss": 1.8438,
+      "step": 15520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6007958734616294,
+      "learning_rate": 9.48599601562426e-06,
+      "loss": 1.3691,
+      "step": 15521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3887834602035176,
+      "learning_rate": 9.482725857077184e-06,
+      "loss": 1.124,
+      "step": 15522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.021898612370989,
+      "learning_rate": 9.479456130369527e-06,
+      "loss": 0.812,
+      "step": 15523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5526898240840343,
+      "learning_rate": 9.476186835592293e-06,
+      "loss": 1.3477,
+      "step": 15524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.183250004389108,
+      "learning_rate": 9.47291797283646e-06,
+      "loss": 0.6187,
+      "step": 15525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.314052576099163,
+      "learning_rate": 9.46964954219299e-06,
+      "loss": 1.0742,
+      "step": 15526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5827566485327842,
+      "learning_rate": 9.46638154375286e-06,
+      "loss": 0.5098,
+      "step": 15527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.34828224028318,
+      "learning_rate": 9.463113977607e-06,
+      "loss": 1.4355,
+      "step": 15528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.713865254755969,
+      "learning_rate": 9.45984684384636e-06,
+      "loss": 1.1035,
+      "step": 15529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5924312502243705,
+      "learning_rate": 9.456580142561854e-06,
+      "loss": 1.1699,
+      "step": 15530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.743742813919437,
+      "learning_rate": 9.453313873844388e-06,
+      "loss": 1.3184,
+      "step": 15531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.420917744359487,
+      "learning_rate": 9.450048037784878e-06,
+      "loss": 0.8018,
+      "step": 15532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.893504965038668,
+      "learning_rate": 9.4467826344742e-06,
+      "loss": 1.3135,
+      "step": 15533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9981873762115128,
+      "learning_rate": 9.443517664003229e-06,
+      "loss": 1.1133,
+      "step": 15534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.716967371820241,
+      "learning_rate": 9.44025312646282e-06,
+      "loss": 0.2269,
+      "step": 15535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.433421121172065,
+      "learning_rate": 9.436989021943831e-06,
+      "loss": 0.9971,
+      "step": 15536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.803310587765386,
+      "learning_rate": 9.433725350537115e-06,
+      "loss": 1.376,
+      "step": 15537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4968251406308477,
+      "learning_rate": 9.430462112333479e-06,
+      "loss": 1.1973,
+      "step": 15538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5831241250881294,
+      "learning_rate": 9.427199307423748e-06,
+      "loss": 1.1074,
+      "step": 15539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.055687929777171,
+      "learning_rate": 9.42393693589871e-06,
+      "loss": 1.7969,
+      "step": 15540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.311787779132914,
+      "learning_rate": 9.420674997849178e-06,
+      "loss": 0.9971,
+      "step": 15541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9484914122778734,
+      "learning_rate": 9.417413493365915e-06,
+      "loss": 1.0498,
+      "step": 15542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5298963153083838,
+      "learning_rate": 9.414152422539693e-06,
+      "loss": 0.6689,
+      "step": 15543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.39534970427279,
+      "learning_rate": 9.410891785461254e-06,
+      "loss": 1.3672,
+      "step": 15544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.05558255362828,
+      "learning_rate": 9.407631582221352e-06,
+      "loss": 1.4883,
+      "step": 15545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5930322780108055,
+      "learning_rate": 9.404371812910723e-06,
+      "loss": 1.2285,
+      "step": 15546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.18268655073647,
+      "learning_rate": 9.401112477620078e-06,
+      "loss": 1.2969,
+      "step": 15547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.650285374372502,
+      "learning_rate": 9.397853576440122e-06,
+      "loss": 1.1543,
+      "step": 15548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.1660904172132645,
+      "learning_rate": 9.394595109461537e-06,
+      "loss": 1.3062,
+      "step": 15549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.017497891305062,
+      "learning_rate": 9.391337076775031e-06,
+      "loss": 0.7188,
+      "step": 15550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6878881839474453,
+      "learning_rate": 9.388079478471256e-06,
+      "loss": 0.7129,
+      "step": 15551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9839064277648246,
+      "learning_rate": 9.384822314640868e-06,
+      "loss": 0.5225,
+      "step": 15552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.194672984549506,
+      "learning_rate": 9.381565585374527e-06,
+      "loss": 1.1465,
+      "step": 15553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.28117725831157,
+      "learning_rate": 9.378309290762846e-06,
+      "loss": 0.7195,
+      "step": 15554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.514660029901645,
+      "learning_rate": 9.37505343089647e-06,
+      "loss": 1.75,
+      "step": 15555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7865851729538176,
+      "learning_rate": 9.371798005865991e-06,
+      "loss": 1.6367,
+      "step": 15556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4901687965999244,
+      "learning_rate": 9.368543015762007e-06,
+      "loss": 1.0547,
+      "step": 15557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.027992307040925,
+      "learning_rate": 9.365288460675115e-06,
+      "loss": 0.5911,
+      "step": 15558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.190811658160536,
+      "learning_rate": 9.36203434069588e-06,
+      "loss": 0.8916,
+      "step": 15559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.577572578836818,
+      "learning_rate": 9.358780655914865e-06,
+      "loss": 1.4492,
+      "step": 15560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5958683601412407,
+      "learning_rate": 9.355527406422604e-06,
+      "loss": 1.082,
+      "step": 15561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.734594560322432,
+      "learning_rate": 9.352274592309656e-06,
+      "loss": 1.3223,
+      "step": 15562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6837969791664444,
+      "learning_rate": 9.34902221366653e-06,
+      "loss": 0.6255,
+      "step": 15563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.01790065775603,
+      "learning_rate": 9.345770270583751e-06,
+      "loss": 1.3428,
+      "step": 15564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.025081676341575,
+      "learning_rate": 9.342518763151815e-06,
+      "loss": 1.0518,
+      "step": 15565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.670665265166032,
+      "learning_rate": 9.339267691461196e-06,
+      "loss": 1.3418,
+      "step": 15566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7151579618692696,
+      "learning_rate": 9.336017055602388e-06,
+      "loss": 0.8379,
+      "step": 15567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8974824961807495,
+      "learning_rate": 9.332766855665853e-06,
+      "loss": 1.21,
+      "step": 15568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.586628429822972,
+      "learning_rate": 9.329517091742038e-06,
+      "loss": 1.5078,
+      "step": 15569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3835561123567395,
+      "learning_rate": 9.326267763921374e-06,
+      "loss": 0.7627,
+      "step": 15570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.991863881827822,
+      "learning_rate": 9.323018872294303e-06,
+      "loss": 1.6094,
+      "step": 15571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.82058219540845,
+      "learning_rate": 9.31977041695123e-06,
+      "loss": 1.4961,
+      "step": 15572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.766314059695261,
+      "learning_rate": 9.316522397982569e-06,
+      "loss": 0.6626,
+      "step": 15573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5393622118782573,
+      "learning_rate": 9.313274815478698e-06,
+      "loss": 1.123,
+      "step": 15574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.346425449729122,
+      "learning_rate": 9.310027669530011e-06,
+      "loss": 1.1992,
+      "step": 15575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1917037171553106,
+      "learning_rate": 9.306780960226865e-06,
+      "loss": 0.3188,
+      "step": 15576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.755550193702298,
+      "learning_rate": 9.303534687659619e-06,
+      "loss": 1.6895,
+      "step": 15577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8875144402160016,
+      "learning_rate": 9.300288851918601e-06,
+      "loss": 1.1211,
+      "step": 15578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.1326766787228735,
+      "learning_rate": 9.297043453094161e-06,
+      "loss": 0.9596,
+      "step": 15579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.946570069294425,
+      "learning_rate": 9.293798491276613e-06,
+      "loss": 1.3428,
+      "step": 15580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.763741493022219,
+      "learning_rate": 9.290553966556256e-06,
+      "loss": 0.9346,
+      "step": 15581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9987622841940036,
+      "learning_rate": 9.287309879023379e-06,
+      "loss": 1.1543,
+      "step": 15582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.094964146163923,
+      "learning_rate": 9.284066228768271e-06,
+      "loss": 1.2324,
+      "step": 15583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.3679985755088495,
+      "learning_rate": 9.28082301588121e-06,
+      "loss": 1.791,
+      "step": 15584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.34568076397023,
+      "learning_rate": 9.277580240452446e-06,
+      "loss": 0.9443,
+      "step": 15585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.519177536371626,
+      "learning_rate": 9.274337902572225e-06,
+      "loss": 0.6309,
+      "step": 15586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.934118719152491,
+      "learning_rate": 9.271096002330768e-06,
+      "loss": 0.9043,
+      "step": 15587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1209473954601235,
+      "learning_rate": 9.267854539818313e-06,
+      "loss": 0.7642,
+      "step": 15588
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1909203084070694,
+      "learning_rate": 9.264613515125064e-06,
+      "loss": 1.0977,
+      "step": 15589
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.37701962580942,
+      "learning_rate": 9.261372928341206e-06,
+      "loss": 1.4492,
+      "step": 15590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.287609194596777,
+      "learning_rate": 9.258132779556933e-06,
+      "loss": 0.7734,
+      "step": 15591
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.554180969978181,
+      "learning_rate": 9.254893068862423e-06,
+      "loss": 0.9209,
+      "step": 15592
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.943871588332084,
+      "learning_rate": 9.251653796347829e-06,
+      "loss": 1.2773,
+      "step": 15593
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5177919520448344,
+      "learning_rate": 9.2484149621033e-06,
+      "loss": 1.375,
+      "step": 15594
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.029209461844455,
+      "learning_rate": 9.24517656621896e-06,
+      "loss": 1.5371,
+      "step": 15595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0598835570177054,
+      "learning_rate": 9.241938608784954e-06,
+      "loss": 0.4644,
+      "step": 15596
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.619680919290837,
+      "learning_rate": 9.238701089891378e-06,
+      "loss": 1.0742,
+      "step": 15597
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.421356145593058,
+      "learning_rate": 9.235464009628337e-06,
+      "loss": 0.7744,
+      "step": 15598
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1955888784615114,
+      "learning_rate": 9.232227368085902e-06,
+      "loss": 1.2344,
+      "step": 15599
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4082720298840568,
+      "learning_rate": 9.228991165354165e-06,
+      "loss": 1.0986,
+      "step": 15600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.896497619148422,
+      "learning_rate": 9.22575540152319e-06,
+      "loss": 1.1279,
+      "step": 15601
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7119279198267943,
+      "learning_rate": 9.22252007668302e-06,
+      "loss": 0.6587,
+      "step": 15602
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.361565589846294,
+      "learning_rate": 9.219285190923693e-06,
+      "loss": 1.3887,
+      "step": 15603
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9210666103858312,
+      "learning_rate": 9.216050744335228e-06,
+      "loss": 0.5776,
+      "step": 15604
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.86245109387813,
+      "learning_rate": 9.212816737007655e-06,
+      "loss": 1.3965,
+      "step": 15605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.939570245278842,
+      "learning_rate": 9.20958316903096e-06,
+      "loss": 1.7344,
+      "step": 15606
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4302132796549105,
+      "learning_rate": 9.20635004049514e-06,
+      "loss": 1.2109,
+      "step": 15607
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6136285809757829,
+      "learning_rate": 9.203117351490162e-06,
+      "loss": 1.0415,
+      "step": 15608
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1085145612832443,
+      "learning_rate": 9.199885102105995e-06,
+      "loss": 1.2422,
+      "step": 15609
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1587035962746888,
+      "learning_rate": 9.196653292432603e-06,
+      "loss": 1.3779,
+      "step": 15610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.833815101202898,
+      "learning_rate": 9.193421922559917e-06,
+      "loss": 0.8906,
+      "step": 15611
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.848231546611903,
+      "learning_rate": 9.190190992577854e-06,
+      "loss": 1.4941,
+      "step": 15612
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.781915392199973,
+      "learning_rate": 9.186960502576348e-06,
+      "loss": 1.0029,
+      "step": 15613
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3781095237977334,
+      "learning_rate": 9.183730452645295e-06,
+      "loss": 1.2856,
+      "step": 15614
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.518707807035251,
+      "learning_rate": 9.180500842874584e-06,
+      "loss": 0.833,
+      "step": 15615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7348706336258852,
+      "learning_rate": 9.177271673354085e-06,
+      "loss": 0.604,
+      "step": 15616
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.03452069515174,
+      "learning_rate": 9.174042944173683e-06,
+      "loss": 1.3203,
+      "step": 15617
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0400315828309505,
+      "learning_rate": 9.170814655423224e-06,
+      "loss": 0.9854,
+      "step": 15618
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.271718741605079,
+      "learning_rate": 9.167586807192538e-06,
+      "loss": 0.6367,
+      "step": 15619
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8939654869441727,
+      "learning_rate": 9.164359399571477e-06,
+      "loss": 0.8926,
+      "step": 15620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.834499025838257,
+      "learning_rate": 9.161132432649839e-06,
+      "loss": 0.6943,
+      "step": 15621
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.620226197706686,
+      "learning_rate": 9.157905906517445e-06,
+      "loss": 1.1895,
+      "step": 15622
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.630177467489784,
+      "learning_rate": 9.154679821264079e-06,
+      "loss": 0.5273,
+      "step": 15623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.715464130022116,
+      "learning_rate": 9.151454176979524e-06,
+      "loss": 1.1855,
+      "step": 15624
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.859511787476399,
+      "learning_rate": 9.14822897375354e-06,
+      "loss": 1.2764,
+      "step": 15625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.877795042007427,
+      "learning_rate": 9.145004211675896e-06,
+      "loss": 0.9175,
+      "step": 15626
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2746176401821465,
+      "learning_rate": 9.141779890836333e-06,
+      "loss": 0.3096,
+      "step": 15627
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8325029524202665,
+      "learning_rate": 9.13855601132457e-06,
+      "loss": 1.2949,
+      "step": 15628
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3448634236790866,
+      "learning_rate": 9.135332573230334e-06,
+      "loss": 1.1113,
+      "step": 15629
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4042283804695073,
+      "learning_rate": 9.132109576643347e-06,
+      "loss": 0.6582,
+      "step": 15630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.424781777699616,
+      "learning_rate": 9.128887021653287e-06,
+      "loss": 1.0432,
+      "step": 15631
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3697871169720806,
+      "learning_rate": 9.125664908349843e-06,
+      "loss": 0.8369,
+      "step": 15632
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.196422918478634,
+      "learning_rate": 9.12244323682267e-06,
+      "loss": 0.6028,
+      "step": 15633
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9109242904289654,
+      "learning_rate": 9.11922200716145e-06,
+      "loss": 1.5234,
+      "step": 15634
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.209360861630047,
+      "learning_rate": 9.116001219455813e-06,
+      "loss": 1.1201,
+      "step": 15635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.140926456598394,
+      "learning_rate": 9.1127808737954e-06,
+      "loss": 1.3574,
+      "step": 15636
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.847438995889006,
+      "learning_rate": 9.109560970269814e-06,
+      "loss": 0.7964,
+      "step": 15637
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1342952730136586,
+      "learning_rate": 9.106341508968683e-06,
+      "loss": 0.9287,
+      "step": 15638
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1699927427460572,
+      "learning_rate": 9.103122489981602e-06,
+      "loss": 1.3789,
+      "step": 15639
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6068452499185977,
+      "learning_rate": 9.09990391339815e-06,
+      "loss": 0.9355,
+      "step": 15640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4710311949264216,
+      "learning_rate": 9.0966857793079e-06,
+      "loss": 1.2012,
+      "step": 15641
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.19110561715128,
+      "learning_rate": 9.093468087800406e-06,
+      "loss": 0.7568,
+      "step": 15642
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.458615518527518,
+      "learning_rate": 9.090250838965223e-06,
+      "loss": 0.9248,
+      "step": 15643
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4549229344471706,
+      "learning_rate": 9.087034032891883e-06,
+      "loss": 1.1719,
+      "step": 15644
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.906486809201432,
+      "learning_rate": 9.083817669669909e-06,
+      "loss": 0.6506,
+      "step": 15645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.29916759201339,
+      "learning_rate": 9.0806017493888e-06,
+      "loss": 2.4512,
+      "step": 15646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.456634829128111,
+      "learning_rate": 9.077386272138066e-06,
+      "loss": 1.043,
+      "step": 15647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8642326820881825,
+      "learning_rate": 9.074171238007198e-06,
+      "loss": 1.0254,
+      "step": 15648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8409506676010334,
+      "learning_rate": 9.07095664708566e-06,
+      "loss": 0.5901,
+      "step": 15649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9601690659053035,
+      "learning_rate": 9.067742499462902e-06,
+      "loss": 0.6026,
+      "step": 15650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.907889939168164,
+      "learning_rate": 9.064528795228396e-06,
+      "loss": 1.1392,
+      "step": 15651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3397516605999287,
+      "learning_rate": 9.061315534471568e-06,
+      "loss": 1.2832,
+      "step": 15652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0998380598571194,
+      "learning_rate": 9.058102717281838e-06,
+      "loss": 0.9209,
+      "step": 15653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.26104047746428,
+      "learning_rate": 9.054890343748613e-06,
+      "loss": 1.4053,
+      "step": 15654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.207492530349203,
+      "learning_rate": 9.051678413961303e-06,
+      "loss": 0.9111,
+      "step": 15655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8539931112167483,
+      "learning_rate": 9.048466928009284e-06,
+      "loss": 1.1851,
+      "step": 15656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3804786133569658,
+      "learning_rate": 9.045255885981946e-06,
+      "loss": 0.6387,
+      "step": 15657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.106614959151819,
+      "learning_rate": 9.04204528796864e-06,
+      "loss": 0.7471,
+      "step": 15658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.736590724763358,
+      "learning_rate": 9.03883513405871e-06,
+      "loss": 0.8096,
+      "step": 15659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5618142721361568,
+      "learning_rate": 9.035625424341506e-06,
+      "loss": 0.5059,
+      "step": 15660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5694543397271863,
+      "learning_rate": 9.032416158906351e-06,
+      "loss": 0.8223,
+      "step": 15661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3700284622994245,
+      "learning_rate": 9.029207337842551e-06,
+      "loss": 0.6963,
+      "step": 15662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.034941409984392,
+      "learning_rate": 9.025998961239402e-06,
+      "loss": 1.1758,
+      "step": 15663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.335083583567181,
+      "learning_rate": 9.022791029186206e-06,
+      "loss": 0.4107,
+      "step": 15664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1127890755864582,
+      "learning_rate": 9.019583541772225e-06,
+      "loss": 0.9453,
+      "step": 15665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9396215042098013,
+      "learning_rate": 9.016376499086737e-06,
+      "loss": 0.8447,
+      "step": 15666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5481317146068707,
+      "learning_rate": 9.013169901218974e-06,
+      "loss": 0.6416,
+      "step": 15667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9339768404680417,
+      "learning_rate": 9.009963748258193e-06,
+      "loss": 0.906,
+      "step": 15668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5751594347848124,
+      "learning_rate": 9.006758040293611e-06,
+      "loss": 0.4937,
+      "step": 15669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.580082919902681,
+      "learning_rate": 9.003552777414442e-06,
+      "loss": 0.8228,
+      "step": 15670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4089183132119243,
+      "learning_rate": 9.000347959709882e-06,
+      "loss": 0.9268,
+      "step": 15671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.947187964931108,
+      "learning_rate": 8.99714358726913e-06,
+      "loss": 0.8477,
+      "step": 15672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2228663520526815,
+      "learning_rate": 8.993939660181357e-06,
+      "loss": 0.9111,
+      "step": 15673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.829524415833835,
+      "learning_rate": 8.99073617853572e-06,
+      "loss": 0.998,
+      "step": 15674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3195625917213016,
+      "learning_rate": 8.987533142421384e-06,
+      "loss": 1.1719,
+      "step": 15675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3266602074862273,
+      "learning_rate": 8.984330551927475e-06,
+      "loss": 0.7969,
+      "step": 15676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4460211822956874,
+      "learning_rate": 8.981128407143136e-06,
+      "loss": 1.1152,
+      "step": 15677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.996635831127405,
+      "learning_rate": 8.977926708157473e-06,
+      "loss": 0.7617,
+      "step": 15678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5721297366000777,
+      "learning_rate": 8.974725455059585e-06,
+      "loss": 1.2871,
+      "step": 15679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.742566566462626,
+      "learning_rate": 8.971524647938555e-06,
+      "loss": 0.8896,
+      "step": 15680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.381135544316559,
+      "learning_rate": 8.968324286883479e-06,
+      "loss": 1.1794,
+      "step": 15681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9771900010217482,
+      "learning_rate": 8.96512437198341e-06,
+      "loss": 0.8243,
+      "step": 15682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.834538517796059,
+      "learning_rate": 8.961924903327404e-06,
+      "loss": 1.3145,
+      "step": 15683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.399417341553773,
+      "learning_rate": 8.958725881004487e-06,
+      "loss": 0.917,
+      "step": 15684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.514589274182659,
+      "learning_rate": 8.9555273051037e-06,
+      "loss": 0.5063,
+      "step": 15685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.788031355992796,
+      "learning_rate": 8.952329175714066e-06,
+      "loss": 0.8691,
+      "step": 15686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.59086612592307,
+      "learning_rate": 8.949131492924578e-06,
+      "loss": 0.876,
+      "step": 15687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9296349275045357,
+      "learning_rate": 8.945934256824218e-06,
+      "loss": 0.6233,
+      "step": 15688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2523646921898868,
+      "learning_rate": 8.94273746750198e-06,
+      "loss": 0.6216,
+      "step": 15689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.062320093796486,
+      "learning_rate": 8.939541125046819e-06,
+      "loss": 0.924,
+      "step": 15690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3411618492751525,
+      "learning_rate": 8.936345229547693e-06,
+      "loss": 1.3301,
+      "step": 15691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8433865981083428,
+      "learning_rate": 8.933149781093531e-06,
+      "loss": 0.8438,
+      "step": 15692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.990873673984251,
+      "learning_rate": 8.92995477977327e-06,
+      "loss": 1.2324,
+      "step": 15693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0381468347119136,
+      "learning_rate": 8.926760225675831e-06,
+      "loss": 1.0215,
+      "step": 15694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.721045916802518,
+      "learning_rate": 8.923566118890114e-06,
+      "loss": 1.375,
+      "step": 15695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5913888887289063,
+      "learning_rate": 8.920372459505008e-06,
+      "loss": 1.0596,
+      "step": 15696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.104318392983308,
+      "learning_rate": 8.917179247609381e-06,
+      "loss": 1.3242,
+      "step": 15697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.719182473348241,
+      "learning_rate": 8.913986483292117e-06,
+      "loss": 1.0957,
+      "step": 15698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0713918499439945,
+      "learning_rate": 8.910794166642058e-06,
+      "loss": 1.1372,
+      "step": 15699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.566848242822941,
+      "learning_rate": 8.907602297748049e-06,
+      "loss": 0.4937,
+      "step": 15700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1854268740793508,
+      "learning_rate": 8.90441087669891e-06,
+      "loss": 1.0562,
+      "step": 15701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6516953948082664,
+      "learning_rate": 8.901219903583463e-06,
+      "loss": 1.2734,
+      "step": 15702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0219751913961708,
+      "learning_rate": 8.898029378490519e-06,
+      "loss": 0.6484,
+      "step": 15703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5732163371863326,
+      "learning_rate": 8.894839301508864e-06,
+      "loss": 1.5625,
+      "step": 15704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7306322040668016,
+      "learning_rate": 8.891649672727276e-06,
+      "loss": 1.2734,
+      "step": 15705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6354347097711512,
+      "learning_rate": 8.888460492234507e-06,
+      "loss": 1.3516,
+      "step": 15706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.021945121053083,
+      "learning_rate": 8.885271760119335e-06,
+      "loss": 0.627,
+      "step": 15707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.834023178611582,
+      "learning_rate": 8.882083476470491e-06,
+      "loss": 0.957,
+      "step": 15708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.798394395099358,
+      "learning_rate": 8.878895641376692e-06,
+      "loss": 1.1914,
+      "step": 15709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.214249357900691,
+      "learning_rate": 8.875708254926674e-06,
+      "loss": 1.209,
+      "step": 15710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.893198329577305,
+      "learning_rate": 8.87252131720912e-06,
+      "loss": 1.2363,
+      "step": 15711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.627875070560739,
+      "learning_rate": 8.869334828312741e-06,
+      "loss": 1.2275,
+      "step": 15712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.681375393378581,
+      "learning_rate": 8.86614878832621e-06,
+      "loss": 0.9785,
+      "step": 15713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.655289988171827,
+      "learning_rate": 8.86296319733818e-06,
+      "loss": 0.7363,
+      "step": 15714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.786355208630682,
+      "learning_rate": 8.859778055437323e-06,
+      "loss": 0.9277,
+      "step": 15715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.718062896307876,
+      "learning_rate": 8.856593362712273e-06,
+      "loss": 0.7578,
+      "step": 15716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9331939403166425,
+      "learning_rate": 8.853409119251656e-06,
+      "loss": 0.5791,
+      "step": 15717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.704928781582549,
+      "learning_rate": 8.850225325144082e-06,
+      "loss": 1.1094,
+      "step": 15718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.6495628725726315,
+      "learning_rate": 8.84704198047817e-06,
+      "loss": 0.9746,
+      "step": 15719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1395742245944236,
+      "learning_rate": 8.843859085342506e-06,
+      "loss": 0.7119,
+      "step": 15720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.300007558893139,
+      "learning_rate": 8.84067663982566e-06,
+      "loss": 0.8059,
+      "step": 15721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0073852084393358,
+      "learning_rate": 8.837494644016209e-06,
+      "loss": 0.915,
+      "step": 15722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.249155981986584,
+      "learning_rate": 8.834313098002698e-06,
+      "loss": 1.6719,
+      "step": 15723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.375346413247658,
+      "learning_rate": 8.831132001873676e-06,
+      "loss": 0.8809,
+      "step": 15724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0181281418743517,
+      "learning_rate": 8.82795135571767e-06,
+      "loss": 0.9126,
+      "step": 15725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.253411673457485,
+      "learning_rate": 8.824771159623186e-06,
+      "loss": 1.0791,
+      "step": 15726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.837647146031095,
+      "learning_rate": 8.821591413678746e-06,
+      "loss": 1.1875,
+      "step": 15727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.682520548541329,
+      "learning_rate": 8.818412117972827e-06,
+      "loss": 1.0283,
+      "step": 15728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.285797745965594,
+      "learning_rate": 8.815233272593915e-06,
+      "loss": 1.0889,
+      "step": 15729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9615039937760286,
+      "learning_rate": 8.81205487763046e-06,
+      "loss": 1.5488,
+      "step": 15730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.87452091824089,
+      "learning_rate": 8.808876933170928e-06,
+      "loss": 0.4438,
+      "step": 15731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.677226754806283,
+      "learning_rate": 8.805699439303772e-06,
+      "loss": 0.542,
+      "step": 15732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.67076623319988,
+      "learning_rate": 8.802522396117405e-06,
+      "loss": 1.4307,
+      "step": 15733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2931610449488726,
+      "learning_rate": 8.799345803700246e-06,
+      "loss": 1.0801,
+      "step": 15734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3894796238753973,
+      "learning_rate": 8.796169662140689e-06,
+      "loss": 0.8633,
+      "step": 15735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8144662553638455,
+      "learning_rate": 8.792993971527142e-06,
+      "loss": 0.739,
+      "step": 15736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.901681077349128,
+      "learning_rate": 8.789818731947978e-06,
+      "loss": 1.2129,
+      "step": 15737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.306740780431709,
+      "learning_rate": 8.786643943491554e-06,
+      "loss": 0.7856,
+      "step": 15738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4605116702824126,
+      "learning_rate": 8.783469606246226e-06,
+      "loss": 0.7969,
+      "step": 15739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.330667458203749,
+      "learning_rate": 8.780295720300333e-06,
+      "loss": 0.7798,
+      "step": 15740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.119577356958369,
+      "learning_rate": 8.777122285742217e-06,
+      "loss": 0.624,
+      "step": 15741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.850512167435743,
+      "learning_rate": 8.773949302660185e-06,
+      "loss": 1.1514,
+      "step": 15742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2103884336569766,
+      "learning_rate": 8.770776771142536e-06,
+      "loss": 0.8403,
+      "step": 15743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6255370907594533,
+      "learning_rate": 8.767604691277554e-06,
+      "loss": 1.04,
+      "step": 15744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.559132666287957,
+      "learning_rate": 8.764433063153535e-06,
+      "loss": 1.0391,
+      "step": 15745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2757350178217255,
+      "learning_rate": 8.761261886858732e-06,
+      "loss": 0.9639,
+      "step": 15746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.439187175098458,
+      "learning_rate": 8.758091162481394e-06,
+      "loss": 1.4961,
+      "step": 15747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.843741068015537,
+      "learning_rate": 8.754920890109766e-06,
+      "loss": 0.6005,
+      "step": 15748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.085892351783466,
+      "learning_rate": 8.751751069832084e-06,
+      "loss": 1.0742,
+      "step": 15749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.311855169919155,
+      "learning_rate": 8.748581701736559e-06,
+      "loss": 1.4707,
+      "step": 15750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4734334823951696,
+      "learning_rate": 8.745412785911384e-06,
+      "loss": 1.1172,
+      "step": 15751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8786483238661287,
+      "learning_rate": 8.742244322444748e-06,
+      "loss": 0.9727,
+      "step": 15752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0784915227985556,
+      "learning_rate": 8.739076311424843e-06,
+      "loss": 0.7593,
+      "step": 15753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8258739860857252,
+      "learning_rate": 8.735908752939822e-06,
+      "loss": 0.9414,
+      "step": 15754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.867487474140338,
+      "learning_rate": 8.732741647077841e-06,
+      "loss": 1.3154,
+      "step": 15755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0156120339544765,
+      "learning_rate": 8.729574993927028e-06,
+      "loss": 0.7637,
+      "step": 15756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1875280105977293,
+      "learning_rate": 8.72640879357553e-06,
+      "loss": 0.9883,
+      "step": 15757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7119754664744886,
+      "learning_rate": 8.72324304611144e-06,
+      "loss": 1.0601,
+      "step": 15758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9123377288591417,
+      "learning_rate": 8.72007775162288e-06,
+      "loss": 0.3794,
+      "step": 15759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.174039571645841,
+      "learning_rate": 8.716912910197924e-06,
+      "loss": 1.1621,
+      "step": 15760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.259193068101473,
+      "learning_rate": 8.71374852192465e-06,
+      "loss": 1.7559,
+      "step": 15761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.126284774915708,
+      "learning_rate": 8.71058458689113e-06,
+      "loss": 0.9961,
+      "step": 15762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9396384343707609,
+      "learning_rate": 8.70742110518541e-06,
+      "loss": 0.2428,
+      "step": 15763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.636124697981741,
+      "learning_rate": 8.704258076895527e-06,
+      "loss": 0.8379,
+      "step": 15764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.375157401239771,
+      "learning_rate": 8.701095502109501e-06,
+      "loss": 0.9375,
+      "step": 15765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.755786270264207,
+      "learning_rate": 8.69793338091536e-06,
+      "loss": 1.228,
+      "step": 15766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9946400823011563,
+      "learning_rate": 8.694771713401087e-06,
+      "loss": 0.8457,
+      "step": 15767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1878056099185876,
+      "learning_rate": 8.69161049965469e-06,
+      "loss": 0.5762,
+      "step": 15768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.000023603369691,
+      "learning_rate": 8.688449739764123e-06,
+      "loss": 0.5215,
+      "step": 15769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7833905659461045,
+      "learning_rate": 8.685289433817368e-06,
+      "loss": 0.8567,
+      "step": 15770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5600490478943,
+      "learning_rate": 8.682129581902369e-06,
+      "loss": 1.1445,
+      "step": 15771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.57062298624836,
+      "learning_rate": 8.678970184107057e-06,
+      "loss": 1.0312,
+      "step": 15772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.822686084514584,
+      "learning_rate": 8.675811240519352e-06,
+      "loss": 0.9912,
+      "step": 15773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.476443122369274,
+      "learning_rate": 8.672652751227186e-06,
+      "loss": 0.624,
+      "step": 15774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.42048890842039,
+      "learning_rate": 8.669494716318444e-06,
+      "loss": 0.4878,
+      "step": 15775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.109693654663509,
+      "learning_rate": 8.666337135881008e-06,
+      "loss": 1.0205,
+      "step": 15776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.1933943982876025,
+      "learning_rate": 8.663180010002772e-06,
+      "loss": 1.0156,
+      "step": 15777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.565800927759245,
+      "learning_rate": 8.660023338771576e-06,
+      "loss": 0.6089,
+      "step": 15778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.187974731689602,
+      "learning_rate": 8.656867122275284e-06,
+      "loss": 1.2246,
+      "step": 15779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.277633810441861,
+      "learning_rate": 8.65371136060173e-06,
+      "loss": 1.2246,
+      "step": 15780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.86444859043305,
+      "learning_rate": 8.650556053838732e-06,
+      "loss": 1.0234,
+      "step": 15781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.197251302036651,
+      "learning_rate": 8.647401202074093e-06,
+      "loss": 1.1768,
+      "step": 15782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0483358940112364,
+      "learning_rate": 8.644246805395633e-06,
+      "loss": 0.9023,
+      "step": 15783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8337730365720426,
+      "learning_rate": 8.641092863891126e-06,
+      "loss": 0.6982,
+      "step": 15784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.689207022826821,
+      "learning_rate": 8.637939377648344e-06,
+      "loss": 1.0303,
+      "step": 15785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.712360107752283,
+      "learning_rate": 8.63478634675503e-06,
+      "loss": 0.7859,
+      "step": 15786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.540318011638948,
+      "learning_rate": 8.63163377129897e-06,
+      "loss": 0.7607,
+      "step": 15787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.155889056123188,
+      "learning_rate": 8.628481651367876e-06,
+      "loss": 1.3838,
+      "step": 15788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3354917375226094,
+      "learning_rate": 8.625329987049472e-06,
+      "loss": 1.1699,
+      "step": 15789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1220567178795444,
+      "learning_rate": 8.622178778431461e-06,
+      "loss": 0.626,
+      "step": 15790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3238398908329114,
+      "learning_rate": 8.619028025601555e-06,
+      "loss": 0.4766,
+      "step": 15791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.049082587176701,
+      "learning_rate": 8.615877728647432e-06,
+      "loss": 1.4253,
+      "step": 15792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8104626574416556,
+      "learning_rate": 8.61272788765676e-06,
+      "loss": 1.0859,
+      "step": 15793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.732545208507437,
+      "learning_rate": 8.609578502717191e-06,
+      "loss": 2.2461,
+      "step": 15794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0965468775030605,
+      "learning_rate": 8.606429573916383e-06,
+      "loss": 1.3008,
+      "step": 15795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.154587109691281,
+      "learning_rate": 8.603281101341974e-06,
+      "loss": 1.2578,
+      "step": 15796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.252900161847326,
+      "learning_rate": 8.600133085081574e-06,
+      "loss": 0.5981,
+      "step": 15797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.002442091285244,
+      "learning_rate": 8.596985525222794e-06,
+      "loss": 0.9043,
+      "step": 15798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.283081573566222,
+      "learning_rate": 8.593838421853222e-06,
+      "loss": 0.8135,
+      "step": 15799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.379127735045945,
+      "learning_rate": 8.590691775060455e-06,
+      "loss": 1.0791,
+      "step": 15800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1644243198313244,
+      "learning_rate": 8.587545584932058e-06,
+      "loss": 0.9184,
+      "step": 15801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.984164549714163,
+      "learning_rate": 8.584399851555582e-06,
+      "loss": 0.9197,
+      "step": 15802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.35504420272008,
+      "learning_rate": 8.581254575018572e-06,
+      "loss": 0.7681,
+      "step": 15803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.414651638988962,
+      "learning_rate": 8.57810975540856e-06,
+      "loss": 1.0605,
+      "step": 15804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9131683129358552,
+      "learning_rate": 8.574965392813079e-06,
+      "loss": 0.8193,
+      "step": 15805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2249380510257133,
+      "learning_rate": 8.57182148731962e-06,
+      "loss": 0.9746,
+      "step": 15806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.149017006148513,
+      "learning_rate": 8.568678039015675e-06,
+      "loss": 0.875,
+      "step": 15807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6160984791175976,
+      "learning_rate": 8.56553504798874e-06,
+      "loss": 0.7349,
+      "step": 15808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.382211398311923,
+      "learning_rate": 8.562392514326276e-06,
+      "loss": 1.0049,
+      "step": 15809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.252549959986418,
+      "learning_rate": 8.559250438115731e-06,
+      "loss": 1.4492,
+      "step": 15810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3284807328498442,
+      "learning_rate": 8.55610881944455e-06,
+      "loss": 1.2207,
+      "step": 15811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6407445395233635,
+      "learning_rate": 8.552967658400174e-06,
+      "loss": 1.1211,
+      "step": 15812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7330431501134553,
+      "learning_rate": 8.549826955070003e-06,
+      "loss": 0.7993,
+      "step": 15813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7188198913710866,
+      "learning_rate": 8.546686709541457e-06,
+      "loss": 0.7354,
+      "step": 15814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3261470713281933,
+      "learning_rate": 8.543546921901926e-06,
+      "loss": 1.5352,
+      "step": 15815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6818099888763127,
+      "learning_rate": 8.540407592238775e-06,
+      "loss": 1.2031,
+      "step": 15816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1080755681819787,
+      "learning_rate": 8.537268720639388e-06,
+      "loss": 0.8579,
+      "step": 15817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.122482053428329,
+      "learning_rate": 8.53413030719111e-06,
+      "loss": 1.4746,
+      "step": 15818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.451732556092529,
+      "learning_rate": 8.530992351981282e-06,
+      "loss": 1.6426,
+      "step": 15819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.558201417618055,
+      "learning_rate": 8.527854855097225e-06,
+      "loss": 1.166,
+      "step": 15820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.83978439083379,
+      "learning_rate": 8.524717816626268e-06,
+      "loss": 1.2292,
+      "step": 15821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.466186125779301,
+      "learning_rate": 8.52158123665571e-06,
+      "loss": 0.8262,
+      "step": 15822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4556555846521615,
+      "learning_rate": 8.518445115272827e-06,
+      "loss": 0.7312,
+      "step": 15823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6234444536320862,
+      "learning_rate": 8.515309452564907e-06,
+      "loss": 0.8857,
+      "step": 15824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6648398241356945,
+      "learning_rate": 8.512174248619225e-06,
+      "loss": 0.917,
+      "step": 15825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3238050076528642,
+      "learning_rate": 8.509039503523019e-06,
+      "loss": 0.6963,
+      "step": 15826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.521572782235734,
+      "learning_rate": 8.50590521736353e-06,
+      "loss": 1.1484,
+      "step": 15827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.090800242926797,
+      "learning_rate": 8.502771390227979e-06,
+      "loss": 0.6515,
+      "step": 15828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.493897567662491,
+      "learning_rate": 8.499638022203588e-06,
+      "loss": 0.9609,
+      "step": 15829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2080235773845653,
+      "learning_rate": 8.496505113377557e-06,
+      "loss": 1.1523,
+      "step": 15830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.970929681629324,
+      "learning_rate": 8.493372663837069e-06,
+      "loss": 1.3604,
+      "step": 15831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7615026777097136,
+      "learning_rate": 8.490240673669289e-06,
+      "loss": 0.7437,
+      "step": 15832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.918523929773558,
+      "learning_rate": 8.487109142961394e-06,
+      "loss": 1.4238,
+      "step": 15833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.05834319330016,
+      "learning_rate": 8.483978071800535e-06,
+      "loss": 1.3398,
+      "step": 15834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.059256568307474,
+      "learning_rate": 8.480847460273845e-06,
+      "loss": 1.3652,
+      "step": 15835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.469345670683988,
+      "learning_rate": 8.477717308468442e-06,
+      "loss": 1.4727,
+      "step": 15836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4088339864899355,
+      "learning_rate": 8.474587616471432e-06,
+      "loss": 0.9888,
+      "step": 15837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.108606767098841,
+      "learning_rate": 8.471458384369931e-06,
+      "loss": 1.0977,
+      "step": 15838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.63682714517062,
+      "learning_rate": 8.468329612251014e-06,
+      "loss": 1.332,
+      "step": 15839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3883886910212486,
+      "learning_rate": 8.465201300201756e-06,
+      "loss": 1.3281,
+      "step": 15840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0983353692919264,
+      "learning_rate": 8.462073448309202e-06,
+      "loss": 0.8604,
+      "step": 15841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6114390541544776,
+      "learning_rate": 8.458946056660413e-06,
+      "loss": 1.1895,
+      "step": 15842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.581360802865266,
+      "learning_rate": 8.455819125342431e-06,
+      "loss": 0.6123,
+      "step": 15843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.671791497762305,
+      "learning_rate": 8.452692654442267e-06,
+      "loss": 1.2637,
+      "step": 15844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.288601765440464,
+      "learning_rate": 8.449566644046922e-06,
+      "loss": 0.6367,
+      "step": 15845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9799244149548554,
+      "learning_rate": 8.446441094243406e-06,
+      "loss": 1.1123,
+      "step": 15846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.998519110784141,
+      "learning_rate": 8.443316005118698e-06,
+      "loss": 1.0703,
+      "step": 15847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0351108464167535,
+      "learning_rate": 8.440191376759765e-06,
+      "loss": 1.1133,
+      "step": 15848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4239114382778126,
+      "learning_rate": 8.437067209253554e-06,
+      "loss": 0.6021,
+      "step": 15849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.261798251308316,
+      "learning_rate": 8.433943502687022e-06,
+      "loss": 1.3525,
+      "step": 15850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3479109214317213,
+      "learning_rate": 8.430820257147104e-06,
+      "loss": 0.9355,
+      "step": 15851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.054583729735635,
+      "learning_rate": 8.427697472720717e-06,
+      "loss": 1.3027,
+      "step": 15852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.1033981238689705,
+      "learning_rate": 8.42457514949476e-06,
+      "loss": 1.1311,
+      "step": 15853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.344238230397963,
+      "learning_rate": 8.421453287556119e-06,
+      "loss": 0.9307,
+      "step": 15854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.083721000189101,
+      "learning_rate": 8.418331886991696e-06,
+      "loss": 0.3633,
+      "step": 15855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9675407757805807,
+      "learning_rate": 8.415210947888345e-06,
+      "loss": 0.8745,
+      "step": 15856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.060634504801401,
+      "learning_rate": 8.41209047033292e-06,
+      "loss": 0.8047,
+      "step": 15857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.709653747258888,
+      "learning_rate": 8.408970454412258e-06,
+      "loss": 0.7949,
+      "step": 15858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.369038034529795,
+      "learning_rate": 8.4058509002132e-06,
+      "loss": 1.3164,
+      "step": 15859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8725330654428396,
+      "learning_rate": 8.402731807822547e-06,
+      "loss": 1.1812,
+      "step": 15860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.067702564866623,
+      "learning_rate": 8.399613177327122e-06,
+      "loss": 1.4062,
+      "step": 15861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5888059635945522,
+      "learning_rate": 8.396495008813704e-06,
+      "loss": 0.7679,
+      "step": 15862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2716357462033425,
+      "learning_rate": 8.39337730236906e-06,
+      "loss": 0.8162,
+      "step": 15863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6359894824823966,
+      "learning_rate": 8.390260058079977e-06,
+      "loss": 1.0393,
+      "step": 15864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.811214126607785,
+      "learning_rate": 8.387143276033191e-06,
+      "loss": 1.0352,
+      "step": 15865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.2520570138835225,
+      "learning_rate": 8.384026956315438e-06,
+      "loss": 1.1289,
+      "step": 15866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7996404845105536,
+      "learning_rate": 8.38091109901346e-06,
+      "loss": 0.8237,
+      "step": 15867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7025071689394173,
+      "learning_rate": 8.377795704213959e-06,
+      "loss": 0.9395,
+      "step": 15868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2554248137652655,
+      "learning_rate": 8.37468077200363e-06,
+      "loss": 0.7554,
+      "step": 15869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.038321982407266,
+      "learning_rate": 8.371566302469172e-06,
+      "loss": 0.876,
+      "step": 15870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.140705142802335,
+      "learning_rate": 8.368452295697249e-06,
+      "loss": 0.7021,
+      "step": 15871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.632755478609352,
+      "learning_rate": 8.365338751774534e-06,
+      "loss": 1.2129,
+      "step": 15872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.319175744415255,
+      "learning_rate": 8.36222567078767e-06,
+      "loss": 0.5811,
+      "step": 15873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2341311450381327,
+      "learning_rate": 8.359113052823295e-06,
+      "loss": 0.8144,
+      "step": 15874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.03541067184584,
+      "learning_rate": 8.35600089796802e-06,
+      "loss": 0.8125,
+      "step": 15875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.47595551588165,
+      "learning_rate": 8.35288920630847e-06,
+      "loss": 1.166,
+      "step": 15876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.400699654923916,
+      "learning_rate": 8.349777977931238e-06,
+      "loss": 0.9629,
+      "step": 15877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4175136550785923,
+      "learning_rate": 8.3466672129229e-06,
+      "loss": 1.3184,
+      "step": 15878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2781319407499936,
+      "learning_rate": 8.34355691137004e-06,
+      "loss": 1.0059,
+      "step": 15879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0848864932918745,
+      "learning_rate": 8.3404470733592e-06,
+      "loss": 0.834,
+      "step": 15880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.74548923807867,
+      "learning_rate": 8.33733769897695e-06,
+      "loss": 2.0254,
+      "step": 15881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1396037540618944,
+      "learning_rate": 8.334228788309804e-06,
+      "loss": 1.053,
+      "step": 15882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.49377146050647,
+      "learning_rate": 8.331120341444279e-06,
+      "loss": 0.8887,
+      "step": 15883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.343703409367649,
+      "learning_rate": 8.328012358466896e-06,
+      "loss": 1.1846,
+      "step": 15884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6964353504737417,
+      "learning_rate": 8.324904839464142e-06,
+      "loss": 0.8311,
+      "step": 15885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.226500643376955,
+      "learning_rate": 8.321797784522501e-06,
+      "loss": 0.9814,
+      "step": 15886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6608311361400765,
+      "learning_rate": 8.318691193728425e-06,
+      "loss": 0.8838,
+      "step": 15887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.388139231487191,
+      "learning_rate": 8.315585067168386e-06,
+      "loss": 1.0273,
+      "step": 15888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3266071261499874,
+      "learning_rate": 8.31247940492883e-06,
+      "loss": 1.2012,
+      "step": 15889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.902960936254059,
+      "learning_rate": 8.309374207096177e-06,
+      "loss": 1.4434,
+      "step": 15890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.203919936891156,
+      "learning_rate": 8.306269473756848e-06,
+      "loss": 1.1689,
+      "step": 15891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.35843158754618,
+      "learning_rate": 8.303165204997231e-06,
+      "loss": 0.9189,
+      "step": 15892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.042915203827174,
+      "learning_rate": 8.300061400903742e-06,
+      "loss": 1.6738,
+      "step": 15893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1802057996778594,
+      "learning_rate": 8.296958061562743e-06,
+      "loss": 0.7754,
+      "step": 15894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.906904557706805,
+      "learning_rate": 8.293855187060601e-06,
+      "loss": 1.2793,
+      "step": 15895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8377508245847873,
+      "learning_rate": 8.29075277748366e-06,
+      "loss": 0.4453,
+      "step": 15896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9085044730330605,
+      "learning_rate": 8.287650832918267e-06,
+      "loss": 1.4258,
+      "step": 15897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.20239977624625,
+      "learning_rate": 8.284549353450758e-06,
+      "loss": 0.8252,
+      "step": 15898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.319546898992724,
+      "learning_rate": 8.281448339167433e-06,
+      "loss": 0.9062,
+      "step": 15899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.8838956324633,
+      "learning_rate": 8.278347790154595e-06,
+      "loss": 1.3633,
+      "step": 15900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6325718953840784,
+      "learning_rate": 8.275247706498523e-06,
+      "loss": 0.709,
+      "step": 15901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.083852947543899,
+      "learning_rate": 8.272148088285508e-06,
+      "loss": 1.002,
+      "step": 15902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5158140682667036,
+      "learning_rate": 8.2690489356018e-06,
+      "loss": 1.0859,
+      "step": 15903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.247905663960445,
+      "learning_rate": 8.265950248533641e-06,
+      "loss": 1.3594,
+      "step": 15904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4463919794967834,
+      "learning_rate": 8.262852027167285e-06,
+      "loss": 0.6229,
+      "step": 15905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4558080105683606,
+      "learning_rate": 8.259754271588932e-06,
+      "loss": 0.8105,
+      "step": 15906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0490463038844795,
+      "learning_rate": 8.25665698188481e-06,
+      "loss": 0.9795,
+      "step": 15907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6179424961905857,
+      "learning_rate": 8.253560158141111e-06,
+      "loss": 1.2871,
+      "step": 15908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.528965867594278,
+      "learning_rate": 8.250463800444006e-06,
+      "loss": 1.3594,
+      "step": 15909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.123759187944553,
+      "learning_rate": 8.247367908879683e-06,
+      "loss": 1.082,
+      "step": 15910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0239444524734362,
+      "learning_rate": 8.24427248353429e-06,
+      "loss": 0.6775,
+      "step": 15911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7156486080565054,
+      "learning_rate": 8.241177524493974e-06,
+      "loss": 0.8169,
+      "step": 15912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.774842500083303,
+      "learning_rate": 8.238083031844856e-06,
+      "loss": 1.4727,
+      "step": 15913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8516395050460077,
+      "learning_rate": 8.234989005673072e-06,
+      "loss": 1.0137,
+      "step": 15914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8221939869800736,
+      "learning_rate": 8.23189544606471e-06,
+      "loss": 0.9141,
+      "step": 15915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.343086533367044,
+      "learning_rate": 8.22880235310588e-06,
+      "loss": 1.1162,
+      "step": 15916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.923754199781289,
+      "learning_rate": 8.225709726882653e-06,
+      "loss": 0.6658,
+      "step": 15917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4728880377389477,
+      "learning_rate": 8.222617567481086e-06,
+      "loss": 0.8418,
+      "step": 15918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.241779494753235,
+      "learning_rate": 8.219525874987252e-06,
+      "loss": 0.5962,
+      "step": 15919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2898097695774635,
+      "learning_rate": 8.21643464948718e-06,
+      "loss": 0.7236,
+      "step": 15920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3286911746996655,
+      "learning_rate": 8.213343891066903e-06,
+      "loss": 1.6602,
+      "step": 15921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.403715963676671,
+      "learning_rate": 8.21025359981242e-06,
+      "loss": 1.3096,
+      "step": 15922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7689099818869103,
+      "learning_rate": 8.207163775809754e-06,
+      "loss": 0.916,
+      "step": 15923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.89597925783601,
+      "learning_rate": 8.204074419144883e-06,
+      "loss": 1.0527,
+      "step": 15924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.726805736992705,
+      "learning_rate": 8.200985529903773e-06,
+      "loss": 1.2441,
+      "step": 15925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.633534778817225,
+      "learning_rate": 8.1978971081724e-06,
+      "loss": 1.1191,
+      "step": 15926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.737431682270469,
+      "learning_rate": 8.194809154036717e-06,
+      "loss": 1.2388,
+      "step": 15927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.517638876623522,
+      "learning_rate": 8.191721667582655e-06,
+      "loss": 1.2861,
+      "step": 15928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.364678088290809,
+      "learning_rate": 8.188634648896135e-06,
+      "loss": 0.833,
+      "step": 15929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.9972286187822865,
+      "learning_rate": 8.185548098063062e-06,
+      "loss": 1.1416,
+      "step": 15930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3095062053404627,
+      "learning_rate": 8.182462015169348e-06,
+      "loss": 0.6846,
+      "step": 15931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5970751430795738,
+      "learning_rate": 8.17937640030087e-06,
+      "loss": 0.8594,
+      "step": 15932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.240961481993505,
+      "learning_rate": 8.176291253543497e-06,
+      "loss": 1.2314,
+      "step": 15933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.439853571299226,
+      "learning_rate": 8.173206574983083e-06,
+      "loss": 1.2827,
+      "step": 15934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.369055575750768,
+      "learning_rate": 8.170122364705479e-06,
+      "loss": 1.8066,
+      "step": 15935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.244936649983461,
+      "learning_rate": 8.167038622796527e-06,
+      "loss": 0.9844,
+      "step": 15936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0616544217892905,
+      "learning_rate": 8.163955349342036e-06,
+      "loss": 1.082,
+      "step": 15937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.706939904324431,
+      "learning_rate": 8.160872544427816e-06,
+      "loss": 1.0317,
+      "step": 15938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546092468991796,
+      "learning_rate": 8.157790208139649e-06,
+      "loss": 1.0156,
+      "step": 15939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.285378156793654,
+      "learning_rate": 8.154708340563332e-06,
+      "loss": 1.2617,
+      "step": 15940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2283450837872856,
+      "learning_rate": 8.151626941784624e-06,
+      "loss": 1.2852,
+      "step": 15941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9278767776662478,
+      "learning_rate": 8.14854601188928e-06,
+      "loss": 0.6082,
+      "step": 15942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0571834581885144,
+      "learning_rate": 8.145465550963022e-06,
+      "loss": 1.4551,
+      "step": 15943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0907581646166964,
+      "learning_rate": 8.142385559091614e-06,
+      "loss": 0.7437,
+      "step": 15944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.225253747228083,
+      "learning_rate": 8.139306036360753e-06,
+      "loss": 1.3086,
+      "step": 15945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.393270218376047,
+      "learning_rate": 8.13622698285614e-06,
+      "loss": 0.96,
+      "step": 15946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0271444057296235,
+      "learning_rate": 8.13314839866346e-06,
+      "loss": 1.1104,
+      "step": 15947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.764673136470534,
+      "learning_rate": 8.1300702838684e-06,
+      "loss": 0.835,
+      "step": 15948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.231186839687015,
+      "learning_rate": 8.126992638556613e-06,
+      "loss": 1.0166,
+      "step": 15949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.227859790465549,
+      "learning_rate": 8.123915462813755e-06,
+      "loss": 1.1211,
+      "step": 15950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7038919805426063,
+      "learning_rate": 8.120838756725451e-06,
+      "loss": 0.9453,
+      "step": 15951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.38026624018383,
+      "learning_rate": 8.117762520377334e-06,
+      "loss": 1.0479,
+      "step": 15952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.839679410421138,
+      "learning_rate": 8.114686753855019e-06,
+      "loss": 1.4258,
+      "step": 15953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.339112130799916,
+      "learning_rate": 8.111611457244098e-06,
+      "loss": 1.2168,
+      "step": 15954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7890209547522877,
+      "learning_rate": 8.108536630630157e-06,
+      "loss": 1.1582,
+      "step": 15955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7155593387010815,
+      "learning_rate": 8.105462274098754e-06,
+      "loss": 0.7305,
+      "step": 15956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.078504509936634,
+      "learning_rate": 8.102388387735469e-06,
+      "loss": 1.0693,
+      "step": 15957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6008820459846,
+      "learning_rate": 8.099314971625834e-06,
+      "loss": 0.8525,
+      "step": 15958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.031502928025438,
+      "learning_rate": 8.096242025855386e-06,
+      "loss": 0.5832,
+      "step": 15959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4512356814404437,
+      "learning_rate": 8.09316955050963e-06,
+      "loss": 0.791,
+      "step": 15960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9020346012397762,
+      "learning_rate": 8.09009754567409e-06,
+      "loss": 1.1191,
+      "step": 15961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.012825157673203,
+      "learning_rate": 8.087026011434245e-06,
+      "loss": 1.1719,
+      "step": 15962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0855749150870677,
+      "learning_rate": 8.083954947875587e-06,
+      "loss": 1.2158,
+      "step": 15963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7768585167497997,
+      "learning_rate": 8.080884355083571e-06,
+      "loss": 0.5959,
+      "step": 15964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0601799409605195,
+      "learning_rate": 8.077814233143663e-06,
+      "loss": 0.9238,
+      "step": 15965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0074498186692233,
+      "learning_rate": 8.074744582141294e-06,
+      "loss": 0.8916,
+      "step": 15966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.331563434580992,
+      "learning_rate": 8.071675402161896e-06,
+      "loss": 0.9463,
+      "step": 15967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5975683509878007,
+      "learning_rate": 8.06860669329087e-06,
+      "loss": 0.8003,
+      "step": 15968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8819939040185782,
+      "learning_rate": 8.065538455613637e-06,
+      "loss": 0.7705,
+      "step": 15969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8230204170183297,
+      "learning_rate": 8.06247068921557e-06,
+      "loss": 0.9756,
+      "step": 15970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7890596055481858,
+      "learning_rate": 8.059403394182045e-06,
+      "loss": 0.9556,
+      "step": 15971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.221703519804636,
+      "learning_rate": 8.056336570598434e-06,
+      "loss": 0.6074,
+      "step": 15972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.6691666219315175,
+      "learning_rate": 8.053270218550068e-06,
+      "loss": 0.4543,
+      "step": 15973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.496822004283128,
+      "learning_rate": 8.050204338122306e-06,
+      "loss": 1.1252,
+      "step": 15974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8766384967579706,
+      "learning_rate": 8.047138929400452e-06,
+      "loss": 1.3672,
+      "step": 15975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.808816664461471,
+      "learning_rate": 8.044073992469822e-06,
+      "loss": 0.7651,
+      "step": 15976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.282331016124678,
+      "learning_rate": 8.041009527415703e-06,
+      "loss": 1.1328,
+      "step": 15977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.831027149425,
+      "learning_rate": 8.03794553432339e-06,
+      "loss": 0.7773,
+      "step": 15978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.946174149797779,
+      "learning_rate": 8.03488201327815e-06,
+      "loss": 0.8354,
+      "step": 15979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.081868408459475,
+      "learning_rate": 8.03181896436523e-06,
+      "loss": 1.0967,
+      "step": 15980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7933133859954944,
+      "learning_rate": 8.028756387669878e-06,
+      "loss": 0.5969,
+      "step": 15981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3105884590179664,
+      "learning_rate": 8.025694283277335e-06,
+      "loss": 0.9287,
+      "step": 15982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.0965531014230585,
+      "learning_rate": 8.02263265127281e-06,
+      "loss": 1.194,
+      "step": 15983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.119067855874697,
+      "learning_rate": 8.019571491741507e-06,
+      "loss": 0.9775,
+      "step": 15984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1242487084204216,
+      "learning_rate": 8.01651080476861e-06,
+      "loss": 0.642,
+      "step": 15985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.055910611957412,
+      "learning_rate": 8.013450590439309e-06,
+      "loss": 1.0254,
+      "step": 15986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.374757814516015,
+      "learning_rate": 8.010390848838764e-06,
+      "loss": 0.8594,
+      "step": 15987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.899398274312858,
+      "learning_rate": 8.007331580052127e-06,
+      "loss": 0.9766,
+      "step": 15988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.284683465780717,
+      "learning_rate": 8.004272784164524e-06,
+      "loss": 1.5762,
+      "step": 15989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.3468950261546695,
+      "learning_rate": 8.001214461261092e-06,
+      "loss": 0.9922,
+      "step": 15990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.733039458002424,
+      "learning_rate": 7.998156611426949e-06,
+      "loss": 0.9189,
+      "step": 15991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1105852964110485,
+      "learning_rate": 7.995099234747183e-06,
+      "loss": 1.2578,
+      "step": 15992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5253076384820585,
+      "learning_rate": 7.992042331306887e-06,
+      "loss": 1.2598,
+      "step": 15993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8172654611620167,
+      "learning_rate": 7.988985901191117e-06,
+      "loss": 1.25,
+      "step": 15994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.921285565425966,
+      "learning_rate": 7.985929944484954e-06,
+      "loss": 1.2812,
+      "step": 15995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.524870608862856,
+      "learning_rate": 7.982874461273438e-06,
+      "loss": 1.0479,
+      "step": 15996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6463227507727254,
+      "learning_rate": 7.979819451641592e-06,
+      "loss": 0.5396,
+      "step": 15997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.748315671654437,
+      "learning_rate": 7.976764915674437e-06,
+      "loss": 1.3862,
+      "step": 15998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1928357036744286,
+      "learning_rate": 7.973710853456984e-06,
+      "loss": 0.6455,
+      "step": 15999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.584148151699039,
+      "learning_rate": 7.970657265074233e-06,
+      "loss": 0.7061,
+      "step": 16000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1567223828026756,
+      "learning_rate": 7.967604150611161e-06,
+      "loss": 1.0166,
+      "step": 16001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.16108617284206,
+      "learning_rate": 7.96455151015272e-06,
+      "loss": 1.0625,
+      "step": 16002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.371933216419039,
+      "learning_rate": 7.961499343783886e-06,
+      "loss": 1.334,
+      "step": 16003
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.357714086986394,
+      "learning_rate": 7.958447651589587e-06,
+      "loss": 1.8086,
+      "step": 16004
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.260183389700518,
+      "learning_rate": 7.955396433654752e-06,
+      "loss": 1.1992,
+      "step": 16005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1900880068415107,
+      "learning_rate": 7.95234569006429e-06,
+      "loss": 1.0479,
+      "step": 16006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8152263566611304,
+      "learning_rate": 7.949295420903111e-06,
+      "loss": 1.6777,
+      "step": 16007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.979765855237241,
+      "learning_rate": 7.946245626256093e-06,
+      "loss": 1.1055,
+      "step": 16008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5047182382278805,
+      "learning_rate": 7.943196306208126e-06,
+      "loss": 1.041,
+      "step": 16009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4783273659971834,
+      "learning_rate": 7.940147460844063e-06,
+      "loss": 0.9092,
+      "step": 16010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.739221086020344,
+      "learning_rate": 7.93709909024874e-06,
+      "loss": 1.623,
+      "step": 16011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.36434305881314,
+      "learning_rate": 7.934051194507013e-06,
+      "loss": 1.2598,
+      "step": 16012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0928095513172518,
+      "learning_rate": 7.931003773703693e-06,
+      "loss": 0.624,
+      "step": 16013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.002903411431086,
+      "learning_rate": 7.927956827923588e-06,
+      "loss": 1.7578,
+      "step": 16014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.469670127627395,
+      "learning_rate": 7.924910357251489e-06,
+      "loss": 0.8467,
+      "step": 16015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.172993142234566,
+      "learning_rate": 7.921864361772191e-06,
+      "loss": 0.4995,
+      "step": 16016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.081011671430073,
+      "learning_rate": 7.918818841570447e-06,
+      "loss": 1.3796,
+      "step": 16017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.597669940108123,
+      "learning_rate": 7.915773796731027e-06,
+      "loss": 1.0283,
+      "step": 16018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4826905880208208,
+      "learning_rate": 7.91272922733867e-06,
+      "loss": 1.2012,
+      "step": 16019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.834236104556707,
+      "learning_rate": 7.909685133478092e-06,
+      "loss": 0.9863,
+      "step": 16020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.3171341916581865,
+      "learning_rate": 7.906641515234028e-06,
+      "loss": 0.8516,
+      "step": 16021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.204205403651578,
+      "learning_rate": 7.903598372691173e-06,
+      "loss": 1.1348,
+      "step": 16022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6146347232796923,
+      "learning_rate": 7.900555705934206e-06,
+      "loss": 0.6934,
+      "step": 16023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.638440102902703,
+      "learning_rate": 7.897513515047819e-06,
+      "loss": 1.1191,
+      "step": 16024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6156859688164746,
+      "learning_rate": 7.894471800116671e-06,
+      "loss": 1.2266,
+      "step": 16025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.303213184588724,
+      "learning_rate": 7.89143056122541e-06,
+      "loss": 1.5469,
+      "step": 16026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.920338599025631,
+      "learning_rate": 7.888389798458663e-06,
+      "loss": 0.7913,
+      "step": 16027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6582829216941748,
+      "learning_rate": 7.885349511901061e-06,
+      "loss": 0.5991,
+      "step": 16028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.317921933534792,
+      "learning_rate": 7.882309701637221e-06,
+      "loss": 1.0205,
+      "step": 16029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.120725933520742,
+      "learning_rate": 7.879270367751735e-06,
+      "loss": 1.4092,
+      "step": 16030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9990822659892786,
+      "learning_rate": 7.876231510329184e-06,
+      "loss": 1.3809,
+      "step": 16031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0523727780871472,
+      "learning_rate": 7.873193129454134e-06,
+      "loss": 1.1191,
+      "step": 16032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9546477218471405,
+      "learning_rate": 7.870155225211151e-06,
+      "loss": 0.5881,
+      "step": 16033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.871239234196662,
+      "learning_rate": 7.867117797684778e-06,
+      "loss": 0.9883,
+      "step": 16034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.195950887043999,
+      "learning_rate": 7.86408084695954e-06,
+      "loss": 0.5457,
+      "step": 16035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.033121495725267,
+      "learning_rate": 7.861044373119946e-06,
+      "loss": 1.0181,
+      "step": 16036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4898029267789004,
+      "learning_rate": 7.858008376250514e-06,
+      "loss": 1.3066,
+      "step": 16037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9875220836637206,
+      "learning_rate": 7.854972856435735e-06,
+      "loss": 1.1611,
+      "step": 16038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.498687304139109,
+      "learning_rate": 7.851937813760085e-06,
+      "loss": 0.8984,
+      "step": 16039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0163017142392747,
+      "learning_rate": 7.848903248308015e-06,
+      "loss": 1.334,
+      "step": 16040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7769967464638134,
+      "learning_rate": 7.84586916016399e-06,
+      "loss": 0.9297,
+      "step": 16041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.275250890017149,
+      "learning_rate": 7.842835549412447e-06,
+      "loss": 1.3594,
+      "step": 16042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.490623246291631,
+      "learning_rate": 7.839802416137807e-06,
+      "loss": 1.1387,
+      "step": 16043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.367866041170451,
+      "learning_rate": 7.836769760424466e-06,
+      "loss": 0.8252,
+      "step": 16044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.408708900841978,
+      "learning_rate": 7.833737582356843e-06,
+      "loss": 1.1426,
+      "step": 16045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4904068950656884,
+      "learning_rate": 7.830705882019318e-06,
+      "loss": 1.3008,
+      "step": 16046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.415280530462686,
+      "learning_rate": 7.82767465949626e-06,
+      "loss": 1.9668,
+      "step": 16047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.629917623076465,
+      "learning_rate": 7.824643914872023e-06,
+      "loss": 0.8906,
+      "step": 16048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.011485208246068,
+      "learning_rate": 7.821613648230949e-06,
+      "loss": 0.9072,
+      "step": 16049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.19441231839557,
+      "learning_rate": 7.81858385965738e-06,
+      "loss": 0.8843,
+      "step": 16050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.703790661922204,
+      "learning_rate": 7.815554549235629e-06,
+      "loss": 0.8865,
+      "step": 16051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3703514583758176,
+      "learning_rate": 7.812525717049999e-06,
+      "loss": 0.7139,
+      "step": 16052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.196080644795608,
+      "learning_rate": 7.80949736318477e-06,
+      "loss": 1.1855,
+      "step": 16053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4720288485599826,
+      "learning_rate": 7.806469487724233e-06,
+      "loss": 0.9321,
+      "step": 16054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.089907978532579,
+      "learning_rate": 7.80344209075266e-06,
+      "loss": 1.084,
+      "step": 16055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2327835711132753,
+      "learning_rate": 7.800415172354291e-06,
+      "loss": 1.4883,
+      "step": 16056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.573636767423824,
+      "learning_rate": 7.797388732613366e-06,
+      "loss": 1.29,
+      "step": 16057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.716714118667914,
+      "learning_rate": 7.794362771614105e-06,
+      "loss": 1.2188,
+      "step": 16058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.879358553714207,
+      "learning_rate": 7.791337289440729e-06,
+      "loss": 0.7412,
+      "step": 16059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.048998689998488,
+      "learning_rate": 7.78831228617743e-06,
+      "loss": 1.125,
+      "step": 16060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1281713494674466,
+      "learning_rate": 7.785287761908385e-06,
+      "loss": 0.311,
+      "step": 16061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.500761202320992,
+      "learning_rate": 7.782263716717783e-06,
+      "loss": 0.9551,
+      "step": 16062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2980041016783925,
+      "learning_rate": 7.779240150689773e-06,
+      "loss": 1.0713,
+      "step": 16063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3653549302214354,
+      "learning_rate": 7.776217063908489e-06,
+      "loss": 0.8916,
+      "step": 16064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.255962331296206,
+      "learning_rate": 7.77319445645808e-06,
+      "loss": 1.5742,
+      "step": 16065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.174903678953583,
+      "learning_rate": 7.77017232842265e-06,
+      "loss": 0.7603,
+      "step": 16066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.317631040398381,
+      "learning_rate": 7.767150679886317e-06,
+      "loss": 0.9043,
+      "step": 16067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6978072602470164,
+      "learning_rate": 7.764129510933166e-06,
+      "loss": 0.7812,
+      "step": 16068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2228870774802116,
+      "learning_rate": 7.761108821647273e-06,
+      "loss": 1.1875,
+      "step": 16069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9365758229921264,
+      "learning_rate": 7.758088612112696e-06,
+      "loss": 0.6685,
+      "step": 16070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.898845249254993,
+      "learning_rate": 7.755068882413499e-06,
+      "loss": 1.3105,
+      "step": 16071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5808594762037047,
+      "learning_rate": 7.752049632633718e-06,
+      "loss": 1.0664,
+      "step": 16072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.03215862695636,
+      "learning_rate": 7.749030862857364e-06,
+      "loss": 0.647,
+      "step": 16073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.525047426439318,
+      "learning_rate": 7.746012573168469e-06,
+      "loss": 0.7812,
+      "step": 16074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3690424031576525,
+      "learning_rate": 7.74299476365101e-06,
+      "loss": 0.8237,
+      "step": 16075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.578976114985608,
+      "learning_rate": 7.73997743438899e-06,
+      "loss": 0.9062,
+      "step": 16076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0626258045748354,
+      "learning_rate": 7.73696058546637e-06,
+      "loss": 0.9927,
+      "step": 16077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.204378569524096,
+      "learning_rate": 7.733944216967113e-06,
+      "loss": 0.896,
+      "step": 16078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.497133170817021,
+      "learning_rate": 7.730928328975148e-06,
+      "loss": 1.125,
+      "step": 16079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1144842365766072,
+      "learning_rate": 7.727912921574425e-06,
+      "loss": 0.8384,
+      "step": 16080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6294789760562436,
+      "learning_rate": 7.724897994848856e-06,
+      "loss": 1.248,
+      "step": 16081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.1055124858478305,
+      "learning_rate": 7.721883548882334e-06,
+      "loss": 0.6079,
+      "step": 16082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8501322570126495,
+      "learning_rate": 7.71886958375876e-06,
+      "loss": 0.5578,
+      "step": 16083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0522057483760694,
+      "learning_rate": 7.715856099562016e-06,
+      "loss": 1.1738,
+      "step": 16084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7414564587399957,
+      "learning_rate": 7.71284309637596e-06,
+      "loss": 0.7363,
+      "step": 16085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.442575403648482,
+      "learning_rate": 7.709830574284446e-06,
+      "loss": 1.3184,
+      "step": 16086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.101625278537413,
+      "learning_rate": 7.706818533371298e-06,
+      "loss": 0.8818,
+      "step": 16087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8228684781361557,
+      "learning_rate": 7.703806973720359e-06,
+      "loss": 0.8857,
+      "step": 16088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7407462742418365,
+      "learning_rate": 7.700795895415432e-06,
+      "loss": 0.6517,
+      "step": 16089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2857218767789402,
+      "learning_rate": 7.697785298540311e-06,
+      "loss": 0.929,
+      "step": 16090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.316196991858947,
+      "learning_rate": 7.694775183178774e-06,
+      "loss": 1.3145,
+      "step": 16091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.895181230265059,
+      "learning_rate": 7.691765549414598e-06,
+      "loss": 1.1021,
+      "step": 16092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.215274614141179,
+      "learning_rate": 7.68875639733155e-06,
+      "loss": 1.1118,
+      "step": 16093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5722944508576076,
+      "learning_rate": 7.685747727013364e-06,
+      "loss": 0.6836,
+      "step": 16094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.453318788685738,
+      "learning_rate": 7.682739538543768e-06,
+      "loss": 1.1006,
+      "step": 16095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.219035499237147,
+      "learning_rate": 7.679731832006476e-06,
+      "loss": 1.209,
+      "step": 16096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.923074049444461,
+      "learning_rate": 7.6767246074852e-06,
+      "loss": 1.249,
+      "step": 16097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.88689919313764,
+      "learning_rate": 7.673717865063628e-06,
+      "loss": 1.2637,
+      "step": 16098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4173437244355367,
+      "learning_rate": 7.670711604825437e-06,
+      "loss": 0.3691,
+      "step": 16099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9893740066204653,
+      "learning_rate": 7.667705826854276e-06,
+      "loss": 1.0459,
+      "step": 16100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5250563964588695,
+      "learning_rate": 7.664700531233807e-06,
+      "loss": 1.1562,
+      "step": 16101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.74339776352426,
+      "learning_rate": 7.66169571804767e-06,
+      "loss": 0.7437,
+      "step": 16102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.7071370386968665,
+      "learning_rate": 7.658691387379485e-06,
+      "loss": 1.3311,
+      "step": 16103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.9981543563310415,
+      "learning_rate": 7.655687539312853e-06,
+      "loss": 0.9893,
+      "step": 16104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2124634421372456,
+      "learning_rate": 7.652684173931382e-06,
+      "loss": 1.873,
+      "step": 16105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.078022805756787,
+      "learning_rate": 7.64968129131865e-06,
+      "loss": 0.543,
+      "step": 16106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6268271741223277,
+      "learning_rate": 7.64667889155822e-06,
+      "loss": 0.5908,
+      "step": 16107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.681882354221293,
+      "learning_rate": 7.643676974733649e-06,
+      "loss": 0.6387,
+      "step": 16108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4365011271523653,
+      "learning_rate": 7.640675540928487e-06,
+      "loss": 1.0293,
+      "step": 16109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.215714325603995,
+      "learning_rate": 7.637674590226251e-06,
+      "loss": 1.1094,
+      "step": 16110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.465117785703974,
+      "learning_rate": 7.63467412271047e-06,
+      "loss": 1.5195,
+      "step": 16111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3378083390237734,
+      "learning_rate": 7.63167413846464e-06,
+      "loss": 0.6498,
+      "step": 16112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.96479803344904,
+      "learning_rate": 7.628674637572239e-06,
+      "loss": 1.2812,
+      "step": 16113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.477412371673958,
+      "learning_rate": 7.62567562011676e-06,
+      "loss": 1.4229,
+      "step": 16114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.055955966467992,
+      "learning_rate": 7.622677086181654e-06,
+      "loss": 1.1367,
+      "step": 16115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8694570430525403,
+      "learning_rate": 7.6196790358503715e-06,
+      "loss": 1.1514,
+      "step": 16116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.533651248177584,
+      "learning_rate": 7.616681469206335e-06,
+      "loss": 1.5518,
+      "step": 16117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.712655209979903,
+      "learning_rate": 7.613684386332989e-06,
+      "loss": 1.0879,
+      "step": 16118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.330578110376096,
+      "learning_rate": 7.610687787313717e-06,
+      "loss": 0.7559,
+      "step": 16119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9601255727886353,
+      "learning_rate": 7.607691672231934e-06,
+      "loss": 0.751,
+      "step": 16120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5285986684112434,
+      "learning_rate": 7.6046960411710055e-06,
+      "loss": 1.4521,
+      "step": 16121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0832988805822104,
+      "learning_rate": 7.601700894214308e-06,
+      "loss": 1.0615,
+      "step": 16122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5747603542861683,
+      "learning_rate": 7.598706231445196e-06,
+      "loss": 1.0898,
+      "step": 16123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9014258595443407,
+      "learning_rate": 7.595712052947004e-06,
+      "loss": 1.1387,
+      "step": 16124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5977129771247016,
+      "learning_rate": 7.5927183588030525e-06,
+      "loss": 1.0684,
+      "step": 16125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2558814638055464,
+      "learning_rate": 7.5897251490966685e-06,
+      "loss": 0.5615,
+      "step": 16126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3629928034520673,
+      "learning_rate": 7.58673242391115e-06,
+      "loss": 1.0117,
+      "step": 16127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4865299967814747,
+      "learning_rate": 7.583740183329768e-06,
+      "loss": 0.6201,
+      "step": 16128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0522669887727636,
+      "learning_rate": 7.580748427435813e-06,
+      "loss": 0.4268,
+      "step": 16129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.61124436570214,
+      "learning_rate": 7.5777571563125345e-06,
+      "loss": 1.1611,
+      "step": 16130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3211424913299425,
+      "learning_rate": 7.574766370043187e-06,
+      "loss": 1.125,
+      "step": 16131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3695107317953346,
+      "learning_rate": 7.571776068710998e-06,
+      "loss": 1.4141,
+      "step": 16132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.307743770770894,
+      "learning_rate": 7.5687862523991846e-06,
+      "loss": 1.0962,
+      "step": 16133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3685430793669635,
+      "learning_rate": 7.565796921190948e-06,
+      "loss": 0.9707,
+      "step": 16134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.555457964423006,
+      "learning_rate": 7.562808075169492e-06,
+      "loss": 1.2168,
+      "step": 16135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4620789802148253,
+      "learning_rate": 7.55981971441799e-06,
+      "loss": 0.8662,
+      "step": 16136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8910477612619876,
+      "learning_rate": 7.556831839019604e-06,
+      "loss": 1.2949,
+      "step": 16137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.812106634817737,
+      "learning_rate": 7.5538444490574705e-06,
+      "loss": 0.9238,
+      "step": 16138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.913581675376917,
+      "learning_rate": 7.550857544614761e-06,
+      "loss": 1.2246,
+      "step": 16139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.513440812879554,
+      "learning_rate": 7.547871125774583e-06,
+      "loss": 0.9883,
+      "step": 16140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.731130564705236,
+      "learning_rate": 7.544885192620047e-06,
+      "loss": 1.3926,
+      "step": 16141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4999341956061243,
+      "learning_rate": 7.54189974523424e-06,
+      "loss": 0.918,
+      "step": 16142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.571356655000804,
+      "learning_rate": 7.538914783700262e-06,
+      "loss": 1.0273,
+      "step": 16143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.134583421726913,
+      "learning_rate": 7.535930308101183e-06,
+      "loss": 1.2549,
+      "step": 16144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.430676571817501,
+      "learning_rate": 7.532946318520051e-06,
+      "loss": 0.6206,
+      "step": 16145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.377228344019433,
+      "learning_rate": 7.529962815039904e-06,
+      "loss": 1.6406,
+      "step": 16146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.041909754807333,
+      "learning_rate": 7.52697979774378e-06,
+      "loss": 0.585,
+      "step": 16147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3610737729081026,
+      "learning_rate": 7.523997266714702e-06,
+      "loss": 1.1805,
+      "step": 16148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.66413417397678,
+      "learning_rate": 7.521015222035665e-06,
+      "loss": 1.1914,
+      "step": 16149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.2963987272631305,
+      "learning_rate": 7.518033663789661e-06,
+      "loss": 0.7812,
+      "step": 16150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9839683176056673,
+      "learning_rate": 7.515052592059654e-06,
+      "loss": 1.0562,
+      "step": 16151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.983278722280983,
+      "learning_rate": 7.5120720069286225e-06,
+      "loss": 1.0576,
+      "step": 16152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.241828384616672,
+      "learning_rate": 7.50909190847951e-06,
+      "loss": 0.9355,
+      "step": 16153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7083090756626,
+      "learning_rate": 7.506112296795248e-06,
+      "loss": 1.5078,
+      "step": 16154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5783608137650134,
+      "learning_rate": 7.503133171958751e-06,
+      "loss": 0.6387,
+      "step": 16155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.8418254351341,
+      "learning_rate": 7.500154534052933e-06,
+      "loss": 1.4434,
+      "step": 16156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6996067502485523,
+      "learning_rate": 7.4971763831607e-06,
+      "loss": 0.5693,
+      "step": 16157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.840596536451677,
+      "learning_rate": 7.494198719364922e-06,
+      "loss": 1.3926,
+      "step": 16158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2855237467233747,
+      "learning_rate": 7.491221542748461e-06,
+      "loss": 1.043,
+      "step": 16159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.992245666097912,
+      "learning_rate": 7.4882448533941826e-06,
+      "loss": 0.9912,
+      "step": 16160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6259897273575263,
+      "learning_rate": 7.4852686513849215e-06,
+      "loss": 0.6748,
+      "step": 16161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8772686838989685,
+      "learning_rate": 7.482292936803506e-06,
+      "loss": 0.772,
+      "step": 16162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0196659478547616,
+      "learning_rate": 7.479317709732736e-06,
+      "loss": 1.3398,
+      "step": 16163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6188963443766626,
+      "learning_rate": 7.476342970255429e-06,
+      "loss": 1.6289,
+      "step": 16164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3588570537845674,
+      "learning_rate": 7.4733687184543626e-06,
+      "loss": 1.2266,
+      "step": 16165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.769085192601371,
+      "learning_rate": 7.470394954412305e-06,
+      "loss": 0.9043,
+      "step": 16166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7977766057545823,
+      "learning_rate": 7.467421678212025e-06,
+      "loss": 0.6636,
+      "step": 16167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.622763782330444,
+      "learning_rate": 7.464448889936254e-06,
+      "loss": 1.2383,
+      "step": 16168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.489246127436776,
+      "learning_rate": 7.46147658966774e-06,
+      "loss": 0.3961,
+      "step": 16169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5330746514441334,
+      "learning_rate": 7.458504777489195e-06,
+      "loss": 0.4177,
+      "step": 16170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5877946686100146,
+      "learning_rate": 7.455533453483318e-06,
+      "loss": 1.1523,
+      "step": 16171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.482482573264614,
+      "learning_rate": 7.452562617732794e-06,
+      "loss": 1.2305,
+      "step": 16172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8236258088988064,
+      "learning_rate": 7.4495922703203215e-06,
+      "loss": 1.3301,
+      "step": 16173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6522363217469165,
+      "learning_rate": 7.446622411328546e-06,
+      "loss": 1.1108,
+      "step": 16174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2531004842199533,
+      "learning_rate": 7.443653040840118e-06,
+      "loss": 1.0791,
+      "step": 16175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.380053234019714,
+      "learning_rate": 7.4406841589376774e-06,
+      "loss": 1.3431,
+      "step": 16176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8494606086091965,
+      "learning_rate": 7.437715765703857e-06,
+      "loss": 1.1738,
+      "step": 16177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.940036556710955,
+      "learning_rate": 7.434747861221256e-06,
+      "loss": 1.5312,
+      "step": 16178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6301986442362275,
+      "learning_rate": 7.431780445572473e-06,
+      "loss": 1.0151,
+      "step": 16179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.374430071755647,
+      "learning_rate": 7.428813518840078e-06,
+      "loss": 1.2617,
+      "step": 16180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6704580554164346,
+      "learning_rate": 7.42584708110666e-06,
+      "loss": 1.2812,
+      "step": 16181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0229390839986183,
+      "learning_rate": 7.422881132454762e-06,
+      "loss": 1.0898,
+      "step": 16182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7024645577364255,
+      "learning_rate": 7.419915672966926e-06,
+      "loss": 0.6926,
+      "step": 16183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.614987644827889,
+      "learning_rate": 7.416950702725672e-06,
+      "loss": 1.0566,
+      "step": 16184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.07445656364901,
+      "learning_rate": 7.413986221813524e-06,
+      "loss": 1.6484,
+      "step": 16185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.812346283362142,
+      "learning_rate": 7.4110222303129865e-06,
+      "loss": 0.7705,
+      "step": 16186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.726554506304107,
+      "learning_rate": 7.408058728306538e-06,
+      "loss": 0.6946,
+      "step": 16187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7321314735780957,
+      "learning_rate": 7.405095715876656e-06,
+      "loss": 1.209,
+      "step": 16188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.051176603446571,
+      "learning_rate": 7.402133193105787e-06,
+      "loss": 0.6274,
+      "step": 16189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.428070173633621,
+      "learning_rate": 7.399171160076399e-06,
+      "loss": 1.3027,
+      "step": 16190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3750069266770395,
+      "learning_rate": 7.396209616870908e-06,
+      "loss": 1.2559,
+      "step": 16191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7689334886084134,
+      "learning_rate": 7.3932485635717395e-06,
+      "loss": 1.0332,
+      "step": 16192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4287819109993203,
+      "learning_rate": 7.390288000261286e-06,
+      "loss": 0.4319,
+      "step": 16193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.161503900867254,
+      "learning_rate": 7.387327927021951e-06,
+      "loss": 1.1885,
+      "step": 16194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.661433200506001,
+      "learning_rate": 7.384368343936115e-06,
+      "loss": 1.2041,
+      "step": 16195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.409828615999743,
+      "learning_rate": 7.381409251086138e-06,
+      "loss": 1.2969,
+      "step": 16196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.033002654743612,
+      "learning_rate": 7.3784506485543605e-06,
+      "loss": 0.9658,
+      "step": 16197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.605832647850728,
+      "learning_rate": 7.375492536423137e-06,
+      "loss": 0.9795,
+      "step": 16198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.663922075306694,
+      "learning_rate": 7.372534914774781e-06,
+      "loss": 1.291,
+      "step": 16199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.000164743845592,
+      "learning_rate": 7.369577783691603e-06,
+      "loss": 1.0811,
+      "step": 16200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.807162817128457,
+      "learning_rate": 7.366621143255886e-06,
+      "loss": 1.0215,
+      "step": 16201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.18985272996676,
+      "learning_rate": 7.363664993549938e-06,
+      "loss": 1.1719,
+      "step": 16202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3589954197197565,
+      "learning_rate": 7.360709334655999e-06,
+      "loss": 1.4805,
+      "step": 16203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.387923417486295,
+      "learning_rate": 7.35775416665635e-06,
+      "loss": 0.9766,
+      "step": 16204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.133680196567791,
+      "learning_rate": 7.354799489633221e-06,
+      "loss": 1.1816,
+      "step": 16205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.5157774106291,
+      "learning_rate": 7.3518453036688276e-06,
+      "loss": 1.6138,
+      "step": 16206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8022021068909089,
+      "learning_rate": 7.348891608845407e-06,
+      "loss": 1.1025,
+      "step": 16207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5534709868746397,
+      "learning_rate": 7.3459384052451454e-06,
+      "loss": 1.1895,
+      "step": 16208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.1327509202720485,
+      "learning_rate": 7.34298569295023e-06,
+      "loss": 1.7832,
+      "step": 16209
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9310332195751263,
+      "learning_rate": 7.340033472042826e-06,
+      "loss": 0.8149,
+      "step": 16210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7760326867709315,
+      "learning_rate": 7.337081742605112e-06,
+      "loss": 0.2441,
+      "step": 16211
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3227111051270715,
+      "learning_rate": 7.334130504719211e-06,
+      "loss": 0.9141,
+      "step": 16212
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.628503368959668,
+      "learning_rate": 7.3311797584672774e-06,
+      "loss": 0.8203,
+      "step": 16213
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.437224543730532,
+      "learning_rate": 7.3282295039314165e-06,
+      "loss": 0.918,
+      "step": 16214
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.644108928727699,
+      "learning_rate": 7.3252797411937275e-06,
+      "loss": 1.1357,
+      "step": 16215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4450881435152505,
+      "learning_rate": 7.3223304703363135e-06,
+      "loss": 1.249,
+      "step": 16216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.077687969475521,
+      "learning_rate": 7.319381691441249e-06,
+      "loss": 1.3145,
+      "step": 16217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.921572140049551,
+      "learning_rate": 7.316433404590584e-06,
+      "loss": 0.958,
+      "step": 16218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.62357491456532,
+      "learning_rate": 7.313485609866385e-06,
+      "loss": 0.9072,
+      "step": 16219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.918034487187243,
+      "learning_rate": 7.310538307350684e-06,
+      "loss": 1.0967,
+      "step": 16220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.639283265629508,
+      "learning_rate": 7.307591497125493e-06,
+      "loss": 0.9868,
+      "step": 16221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.39635324780063,
+      "learning_rate": 7.304645179272834e-06,
+      "loss": 0.9385,
+      "step": 16222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.922495077502671,
+      "learning_rate": 7.301699353874689e-06,
+      "loss": 0.9297,
+      "step": 16223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.006536356716448,
+      "learning_rate": 7.298754021013055e-06,
+      "loss": 0.7754,
+      "step": 16224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3780201714807245,
+      "learning_rate": 7.295809180769889e-06,
+      "loss": 1.2285,
+      "step": 16225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5960142217573408,
+      "learning_rate": 7.292864833227148e-06,
+      "loss": 0.8701,
+      "step": 16226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9762496183059493,
+      "learning_rate": 7.2899209784667574e-06,
+      "loss": 0.8955,
+      "step": 16227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5554976326145042,
+      "learning_rate": 7.28697761657067e-06,
+      "loss": 0.9453,
+      "step": 16228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3257424881498174,
+      "learning_rate": 7.28403474762078e-06,
+      "loss": 1.2734,
+      "step": 16229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9253936633087774,
+      "learning_rate": 7.281092371698986e-06,
+      "loss": 0.7812,
+      "step": 16230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.778782993480737,
+      "learning_rate": 7.278150488887184e-06,
+      "loss": 1.6836,
+      "step": 16231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.303818076617678,
+      "learning_rate": 7.2752090992672324e-06,
+      "loss": 0.9893,
+      "step": 16232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.653102131487776,
+      "learning_rate": 7.272268202921004e-06,
+      "loss": 0.4023,
+      "step": 16233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4506588237689044,
+      "learning_rate": 7.269327799930334e-06,
+      "loss": 1.0254,
+      "step": 16234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.560073417867753,
+      "learning_rate": 7.266387890377052e-06,
+      "loss": 0.4199,
+      "step": 16235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2198849851008986,
+      "learning_rate": 7.263448474342968e-06,
+      "loss": 0.8103,
+      "step": 16236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.490648144537003,
+      "learning_rate": 7.260509551909903e-06,
+      "loss": 1.752,
+      "step": 16237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.268610493387798,
+      "learning_rate": 7.257571123159632e-06,
+      "loss": 0.5833,
+      "step": 16238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0586138611347917,
+      "learning_rate": 7.254633188173926e-06,
+      "loss": 1.2207,
+      "step": 16239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.74819724937726,
+      "learning_rate": 7.251695747034557e-06,
+      "loss": 0.9844,
+      "step": 16240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.971393842585412,
+      "learning_rate": 7.248758799823277e-06,
+      "loss": 1.5742,
+      "step": 16241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.8522707583095,
+      "learning_rate": 7.245822346621811e-06,
+      "loss": 0.7861,
+      "step": 16242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.595423239148814,
+      "learning_rate": 7.242886387511883e-06,
+      "loss": 1.0186,
+      "step": 16243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.187517220565588,
+      "learning_rate": 7.239950922575189e-06,
+      "loss": 1.0234,
+      "step": 16244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5354989744521284,
+      "learning_rate": 7.237015951893441e-06,
+      "loss": 1.1514,
+      "step": 16245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4722962802933814,
+      "learning_rate": 7.234081475548307e-06,
+      "loss": 0.9795,
+      "step": 16246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.354566190584954,
+      "learning_rate": 7.2311474936214515e-06,
+      "loss": 1.7031,
+      "step": 16247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.41969910859767,
+      "learning_rate": 7.228214006194519e-06,
+      "loss": 0.9824,
+      "step": 16248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.218238177864701,
+      "learning_rate": 7.2252810133491586e-06,
+      "loss": 0.8135,
+      "step": 16249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3207486445957093,
+      "learning_rate": 7.222348515166999e-06,
+      "loss": 0.9609,
+      "step": 16250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.005515837107297,
+      "learning_rate": 7.21941651172964e-06,
+      "loss": 1.2061,
+      "step": 16251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3568143799854715,
+      "learning_rate": 7.216485003118684e-06,
+      "loss": 0.749,
+      "step": 16252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.507798619692409,
+      "learning_rate": 7.213553989415703e-06,
+      "loss": 1.0059,
+      "step": 16253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.8363380081399905,
+      "learning_rate": 7.21062347070228e-06,
+      "loss": 0.7571,
+      "step": 16254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.833267432489465,
+      "learning_rate": 7.207693447059962e-06,
+      "loss": 1.4062,
+      "step": 16255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4438184372472795,
+      "learning_rate": 7.204763918570295e-06,
+      "loss": 1.0361,
+      "step": 16256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6300278604866594,
+      "learning_rate": 7.201834885314795e-06,
+      "loss": 0.8677,
+      "step": 16257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.51421374888865,
+      "learning_rate": 7.198906347374984e-06,
+      "loss": 1.3398,
+      "step": 16258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7487140596615367,
+      "learning_rate": 7.195978304832371e-06,
+      "loss": 0.8047,
+      "step": 16259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7592158650772496,
+      "learning_rate": 7.193050757768432e-06,
+      "loss": 0.5864,
+      "step": 16260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.678106432554899,
+      "learning_rate": 7.190123706264637e-06,
+      "loss": 1.6484,
+      "step": 16261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6126979854192456,
+      "learning_rate": 7.1871971504024554e-06,
+      "loss": 0.9199,
+      "step": 16262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7848688808336002,
+      "learning_rate": 7.184271090263323e-06,
+      "loss": 0.6405,
+      "step": 16263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.565955243209009,
+      "learning_rate": 7.181345525928676e-06,
+      "loss": 0.8179,
+      "step": 16264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.01663088538561,
+      "learning_rate": 7.17842045747992e-06,
+      "loss": 1.0156,
+      "step": 16265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.017028088487344,
+      "learning_rate": 7.175495884998476e-06,
+      "loss": 1.3633,
+      "step": 16266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6020251988488488,
+      "learning_rate": 7.172571808565717e-06,
+      "loss": 0.6958,
+      "step": 16267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.426083654944189,
+      "learning_rate": 7.169648228263034e-06,
+      "loss": 1.29,
+      "step": 16268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.127913217187579,
+      "learning_rate": 7.166725144171785e-06,
+      "loss": 0.5234,
+      "step": 16269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7656426725523366,
+      "learning_rate": 7.163802556373303e-06,
+      "loss": 0.7773,
+      "step": 16270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.284788822298257,
+      "learning_rate": 7.160880464948946e-06,
+      "loss": 1.5459,
+      "step": 16271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.551936193935675,
+      "learning_rate": 7.157958869980022e-06,
+      "loss": 1.043,
+      "step": 16272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.795476800331795,
+      "learning_rate": 7.15503777154784e-06,
+      "loss": 1.1045,
+      "step": 16273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.351647312837651,
+      "learning_rate": 7.152117169733683e-06,
+      "loss": 0.8174,
+      "step": 16274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0985204525456016,
+      "learning_rate": 7.149197064618849e-06,
+      "loss": 0.6797,
+      "step": 16275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.895543877568397,
+      "learning_rate": 7.146277456284592e-06,
+      "loss": 0.825,
+      "step": 16276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.913891007437076,
+      "learning_rate": 7.143358344812157e-06,
+      "loss": 0.4102,
+      "step": 16277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.489224330977412,
+      "learning_rate": 7.1404397302827915e-06,
+      "loss": 0.9209,
+      "step": 16278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.618153225948527,
+      "learning_rate": 7.137521612777725e-06,
+      "loss": 0.7743,
+      "step": 16279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5605707847090835,
+      "learning_rate": 7.134603992378158e-06,
+      "loss": 1.054,
+      "step": 16280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2647091982268215,
+      "learning_rate": 7.131686869165291e-06,
+      "loss": 0.9492,
+      "step": 16281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.110974169476849,
+      "learning_rate": 7.128770243220298e-06,
+      "loss": 1.1387,
+      "step": 16282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.784505417634773,
+      "learning_rate": 7.125854114624358e-06,
+      "loss": 1.2715,
+      "step": 16283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0141541784949175,
+      "learning_rate": 7.122938483458624e-06,
+      "loss": 0.5386,
+      "step": 16284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2065799595404245,
+      "learning_rate": 7.1200233498042335e-06,
+      "loss": 1.0684,
+      "step": 16285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1628326336590873,
+      "learning_rate": 7.117108713742307e-06,
+      "loss": 0.5278,
+      "step": 16286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.842813765472951,
+      "learning_rate": 7.114194575353963e-06,
+      "loss": 0.8291,
+      "step": 16287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.812029332516279,
+      "learning_rate": 7.111280934720313e-06,
+      "loss": 0.5425,
+      "step": 16288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9357861968514638,
+      "learning_rate": 7.108367791922429e-06,
+      "loss": 0.7578,
+      "step": 16289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0839447927320522,
+      "learning_rate": 7.105455147041387e-06,
+      "loss": 0.657,
+      "step": 16290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.638722720894075,
+      "learning_rate": 7.102543000158238e-06,
+      "loss": 1.0752,
+      "step": 16291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.419210136864885,
+      "learning_rate": 7.099631351354036e-06,
+      "loss": 1.293,
+      "step": 16292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1500447527416022,
+      "learning_rate": 7.096720200709808e-06,
+      "loss": 1.3574,
+      "step": 16293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2735773388985328,
+      "learning_rate": 7.093809548306568e-06,
+      "loss": 1.1094,
+      "step": 16294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.411472510232673,
+      "learning_rate": 7.0908993942253034e-06,
+      "loss": 1.0037,
+      "step": 16295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9984907773978744,
+      "learning_rate": 7.087989738547032e-06,
+      "loss": 0.7539,
+      "step": 16296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.619432630767378,
+      "learning_rate": 7.0850805813527154e-06,
+      "loss": 0.8657,
+      "step": 16297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0146048470860825,
+      "learning_rate": 7.082171922723313e-06,
+      "loss": 1.2617,
+      "step": 16298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.549081607067484,
+      "learning_rate": 7.079263762739763e-06,
+      "loss": 1.457,
+      "step": 16299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.506860193959928,
+      "learning_rate": 7.076356101483014e-06,
+      "loss": 0.5776,
+      "step": 16300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.160967222158443,
+      "learning_rate": 7.07344893903398e-06,
+      "loss": 1.1289,
+      "step": 16301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.334204415042852,
+      "learning_rate": 7.070542275473563e-06,
+      "loss": 1.1914,
+      "step": 16302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0135226654673835,
+      "learning_rate": 7.067636110882647e-06,
+      "loss": 0.916,
+      "step": 16303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.21264128021809,
+      "learning_rate": 7.064730445342127e-06,
+      "loss": 0.7617,
+      "step": 16304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.204504065142702,
+      "learning_rate": 7.061825278932849e-06,
+      "loss": 0.8281,
+      "step": 16305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.437198467486908,
+      "learning_rate": 7.058920611735678e-06,
+      "loss": 1.0615,
+      "step": 16306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1991428836681415,
+      "learning_rate": 7.056016443831445e-06,
+      "loss": 0.8218,
+      "step": 16307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.519592757546226,
+      "learning_rate": 7.053112775300957e-06,
+      "loss": 1.293,
+      "step": 16308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.478341892367037,
+      "learning_rate": 7.0502096062250475e-06,
+      "loss": 0.7246,
+      "step": 16309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.389392011234157,
+      "learning_rate": 7.047306936684495e-06,
+      "loss": 0.9434,
+      "step": 16310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.54656802708729,
+      "learning_rate": 7.0444047667600795e-06,
+      "loss": 1.125,
+      "step": 16311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.808628469366635,
+      "learning_rate": 7.041503096532565e-06,
+      "loss": 0.772,
+      "step": 16312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.784229621696697,
+      "learning_rate": 7.038601926082714e-06,
+      "loss": 1.1719,
+      "step": 16313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.908315273400081,
+      "learning_rate": 7.035701255491253e-06,
+      "loss": 1.0244,
+      "step": 16314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.266991140937199,
+      "learning_rate": 7.032801084838919e-06,
+      "loss": 1.7754,
+      "step": 16315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.074829640952269,
+      "learning_rate": 7.0299014142064106e-06,
+      "loss": 1.5449,
+      "step": 16316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.811841524984657,
+      "learning_rate": 7.0270022436744386e-06,
+      "loss": 1.1836,
+      "step": 16317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.860472241307185,
+      "learning_rate": 7.024103573323673e-06,
+      "loss": 1.8535,
+      "step": 16318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.653349408698839,
+      "learning_rate": 7.02120540323479e-06,
+      "loss": 0.8398,
+      "step": 16319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.311403169069678,
+      "learning_rate": 7.018307733488436e-06,
+      "loss": 0.6738,
+      "step": 16320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.739481734203537,
+      "learning_rate": 7.015410564165262e-06,
+      "loss": 0.8916,
+      "step": 16321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.197283802019737,
+      "learning_rate": 7.0125138953458904e-06,
+      "loss": 0.4749,
+      "step": 16322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.073547254588257,
+      "learning_rate": 7.0096177271109285e-06,
+      "loss": 1.0039,
+      "step": 16323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3232505096501734,
+      "learning_rate": 7.006722059540988e-06,
+      "loss": 1.376,
+      "step": 16324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2969246087703503,
+      "learning_rate": 7.003826892716639e-06,
+      "loss": 0.7769,
+      "step": 16325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7650683040208066,
+      "learning_rate": 7.000932226718471e-06,
+      "loss": 0.9592,
+      "step": 16326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.32909019995263,
+      "learning_rate": 6.998038061627029e-06,
+      "loss": 0.9502,
+      "step": 16327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.641992333954221,
+      "learning_rate": 6.99514439752286e-06,
+      "loss": 1.4414,
+      "step": 16328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8325472657585693,
+      "learning_rate": 6.992251234486486e-06,
+      "loss": 0.7227,
+      "step": 16329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.332588780912187,
+      "learning_rate": 6.989358572598437e-06,
+      "loss": 0.7383,
+      "step": 16330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.381265687465414,
+      "learning_rate": 6.986466411939207e-06,
+      "loss": 1.082,
+      "step": 16331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9825398659739695,
+      "learning_rate": 6.983574752589275e-06,
+      "loss": 1.0146,
+      "step": 16332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.26716441848482,
+      "learning_rate": 6.980683594629123e-06,
+      "loss": 1.2656,
+      "step": 16333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.966100212909615,
+      "learning_rate": 6.977792938139219e-06,
+      "loss": 1.8008,
+      "step": 16334
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4659427181870797,
+      "learning_rate": 6.974902783199999e-06,
+      "loss": 1.127,
+      "step": 16335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.244607397129515,
+      "learning_rate": 6.9720131298918994e-06,
+      "loss": 1.6562,
+      "step": 16336
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.573431655886418,
+      "learning_rate": 6.969123978295325e-06,
+      "loss": 0.8652,
+      "step": 16337
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.543368494166573,
+      "learning_rate": 6.9662353284907e-06,
+      "loss": 0.9658,
+      "step": 16338
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.230482004530451,
+      "learning_rate": 6.963347180558403e-06,
+      "loss": 0.5718,
+      "step": 16339
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.657303275922007,
+      "learning_rate": 6.960459534578811e-06,
+      "loss": 0.9854,
+      "step": 16340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.532811754474848,
+      "learning_rate": 6.957572390632278e-06,
+      "loss": 1.1211,
+      "step": 16341
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.325488698510843,
+      "learning_rate": 6.954685748799161e-06,
+      "loss": 0.7117,
+      "step": 16342
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.488618980495855,
+      "learning_rate": 6.951799609159801e-06,
+      "loss": 0.6328,
+      "step": 16343
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5700109174972847,
+      "learning_rate": 6.94891397179451e-06,
+      "loss": 1.3115,
+      "step": 16344
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.873548058130021,
+      "learning_rate": 6.946028836783597e-06,
+      "loss": 1.0176,
+      "step": 16345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9602275264765283,
+      "learning_rate": 6.943144204207341e-06,
+      "loss": 0.8188,
+      "step": 16346
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5800496064619665,
+      "learning_rate": 6.94026007414604e-06,
+      "loss": 1.5391,
+      "step": 16347
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.678023138773344,
+      "learning_rate": 6.93737644667995e-06,
+      "loss": 1.3008,
+      "step": 16348
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7188677087955764,
+      "learning_rate": 6.934493321889319e-06,
+      "loss": 0.875,
+      "step": 16349
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.27213619782326,
+      "learning_rate": 6.931610699854374e-06,
+      "loss": 0.8247,
+      "step": 16350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.486555435079139,
+      "learning_rate": 6.928728580655353e-06,
+      "loss": 2.0137,
+      "step": 16351
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.028678440000076,
+      "learning_rate": 6.925846964372462e-06,
+      "loss": 0.8059,
+      "step": 16352
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1434898418793566,
+      "learning_rate": 6.922965851085894e-06,
+      "loss": 0.8252,
+      "step": 16353
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.29952839809946,
+      "learning_rate": 6.9200852408758175e-06,
+      "loss": 1.5371,
+      "step": 16354
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.281686081745714,
+      "learning_rate": 6.9172051338224185e-06,
+      "loss": 0.9805,
+      "step": 16355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.884963656745124,
+      "learning_rate": 6.914325530005839e-06,
+      "loss": 0.9658,
+      "step": 16356
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.481415996408555,
+      "learning_rate": 6.911446429506216e-06,
+      "loss": 1.5039,
+      "step": 16357
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.165285313829711,
+      "learning_rate": 6.9085678324036675e-06,
+      "loss": 1.0449,
+      "step": 16358
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.550199994490992,
+      "learning_rate": 6.90568973877832e-06,
+      "loss": 0.9336,
+      "step": 16359
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4991077738309264,
+      "learning_rate": 6.902812148710253e-06,
+      "loss": 1.3975,
+      "step": 16360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6425806512136853,
+      "learning_rate": 6.899935062279564e-06,
+      "loss": 1.0312,
+      "step": 16361
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.085117820332318,
+      "learning_rate": 6.897058479566312e-06,
+      "loss": 1.25,
+      "step": 16362
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.04720278557589,
+      "learning_rate": 6.894182400650548e-06,
+      "loss": 1.1367,
+      "step": 16363
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9557181881640062,
+      "learning_rate": 6.891306825612323e-06,
+      "loss": 1.709,
+      "step": 16364
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.791531746111991,
+      "learning_rate": 6.888431754531657e-06,
+      "loss": 1.0049,
+      "step": 16365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5324973341540873,
+      "learning_rate": 6.885557187488561e-06,
+      "loss": 0.9932,
+      "step": 16366
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.673721895520178,
+      "learning_rate": 6.882683124563027e-06,
+      "loss": 0.5171,
+      "step": 16367
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.486060620972345,
+      "learning_rate": 6.879809565835055e-06,
+      "loss": 0.4861,
+      "step": 16368
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4394456385672356,
+      "learning_rate": 6.876936511384599e-06,
+      "loss": 0.8975,
+      "step": 16369
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.981463021159222,
+      "learning_rate": 6.874063961291627e-06,
+      "loss": 1.1191,
+      "step": 16370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6020957239943794,
+      "learning_rate": 6.871191915636077e-06,
+      "loss": 1.1348,
+      "step": 16371
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6060888350255866,
+      "learning_rate": 6.868320374497869e-06,
+      "loss": 0.9092,
+      "step": 16372
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1797422362169137,
+      "learning_rate": 6.8654493379569305e-06,
+      "loss": 1.0596,
+      "step": 16373
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.799092899289495,
+      "learning_rate": 6.862578806093156e-06,
+      "loss": 1.2852,
+      "step": 16374
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6309335948780537,
+      "learning_rate": 6.859708778986421e-06,
+      "loss": 1.0107,
+      "step": 16375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.0470520555363745,
+      "learning_rate": 6.856839256716613e-06,
+      "loss": 1.373,
+      "step": 16376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4887805679605868,
+      "learning_rate": 6.8539702393635805e-06,
+      "loss": 1.1953,
+      "step": 16377
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.162185571115303,
+      "learning_rate": 6.85110172700717e-06,
+      "loss": 1.0742,
+      "step": 16378
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.130669827514201,
+      "learning_rate": 6.848233719727204e-06,
+      "loss": 0.832,
+      "step": 16379
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.347130670905955,
+      "learning_rate": 6.845366217603502e-06,
+      "loss": 1.0898,
+      "step": 16380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5304818577322408,
+      "learning_rate": 6.842499220715879e-06,
+      "loss": 1.3242,
+      "step": 16381
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3547574767591306,
+      "learning_rate": 6.839632729144105e-06,
+      "loss": 0.4307,
+      "step": 16382
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1897374696238043,
+      "learning_rate": 6.836766742967962e-06,
+      "loss": 1.2812,
+      "step": 16383
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8776494474651466,
+      "learning_rate": 6.8339012622672e-06,
+      "loss": 1.0625,
+      "step": 16384
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.395125816609439,
+      "learning_rate": 6.831036287121576e-06,
+      "loss": 0.7939,
+      "step": 16385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2430377903075014,
+      "learning_rate": 6.828171817610815e-06,
+      "loss": 1.1152,
+      "step": 16386
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2035335946461934,
+      "learning_rate": 6.825307853814636e-06,
+      "loss": 0.802,
+      "step": 16387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.397852608912704,
+      "learning_rate": 6.822444395812733e-06,
+      "loss": 0.8198,
+      "step": 16388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.391565148586194,
+      "learning_rate": 6.819581443684803e-06,
+      "loss": 0.6484,
+      "step": 16389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8139966934821885,
+      "learning_rate": 6.816718997510527e-06,
+      "loss": 1.0454,
+      "step": 16390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.926648885776263,
+      "learning_rate": 6.813857057369558e-06,
+      "loss": 0.8105,
+      "step": 16391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.212802614438562,
+      "learning_rate": 6.810995623341545e-06,
+      "loss": 0.7471,
+      "step": 16392
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7580154819731724,
+      "learning_rate": 6.80813469550611e-06,
+      "loss": 0.8845,
+      "step": 16393
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6556127509891294,
+      "learning_rate": 6.805274273942888e-06,
+      "loss": 0.7748,
+      "step": 16394
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.559802240243428,
+      "learning_rate": 6.8024143587314766e-06,
+      "loss": 1.0938,
+      "step": 16395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7690914188481577,
+      "learning_rate": 6.799554949951459e-06,
+      "loss": 1.2266,
+      "step": 16396
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3920774348556746,
+      "learning_rate": 6.796696047682416e-06,
+      "loss": 1.1836,
+      "step": 16397
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8733459141757294,
+      "learning_rate": 6.793837652003921e-06,
+      "loss": 1.2168,
+      "step": 16398
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6950711791877913,
+      "learning_rate": 6.7909797629955084e-06,
+      "loss": 1.3047,
+      "step": 16399
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9786726713208243,
+      "learning_rate": 6.788122380736719e-06,
+      "loss": 1.2305,
+      "step": 16400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8321536916600705,
+      "learning_rate": 6.785265505307062e-06,
+      "loss": 0.9639,
+      "step": 16401
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.003180827223075,
+      "learning_rate": 6.782409136786058e-06,
+      "loss": 1.041,
+      "step": 16402
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.514022384364953,
+      "learning_rate": 6.779553275253192e-06,
+      "loss": 0.9448,
+      "step": 16403
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.871905648535579,
+      "learning_rate": 6.776697920787941e-06,
+      "loss": 1.4824,
+      "step": 16404
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.930535689906574,
+      "learning_rate": 6.77384307346976e-06,
+      "loss": 1.0532,
+      "step": 16405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.165313246580462,
+      "learning_rate": 6.770988733378106e-06,
+      "loss": 1.1973,
+      "step": 16406
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2677565923225496,
+      "learning_rate": 6.768134900592421e-06,
+      "loss": 1.1055,
+      "step": 16407
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5722445468792856,
+      "learning_rate": 6.765281575192123e-06,
+      "loss": 1.2256,
+      "step": 16408
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.233090865319783,
+      "learning_rate": 6.762428757256611e-06,
+      "loss": 0.8496,
+      "step": 16409
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.421400945425414,
+      "learning_rate": 6.759576446865279e-06,
+      "loss": 1.6016,
+      "step": 16410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.334656659343742,
+      "learning_rate": 6.756724644097515e-06,
+      "loss": 0.7412,
+      "step": 16411
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.3442998939046715,
+      "learning_rate": 6.753873349032675e-06,
+      "loss": 0.6936,
+      "step": 16412
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3751328053659493,
+      "learning_rate": 6.751022561750106e-06,
+      "loss": 0.6727,
+      "step": 16413
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.7497601666397316,
+      "learning_rate": 6.748172282329157e-06,
+      "loss": 1.791,
+      "step": 16414
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.532378690400633,
+      "learning_rate": 6.745322510849142e-06,
+      "loss": 0.7849,
+      "step": 16415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.164677721841017,
+      "learning_rate": 6.7424732473893635e-06,
+      "loss": 1.2891,
+      "step": 16416
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6523195810556939,
+      "learning_rate": 6.739624492029131e-06,
+      "loss": 1.2012,
+      "step": 16417
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.114542745030178,
+      "learning_rate": 6.736776244847707e-06,
+      "loss": 0.8564,
+      "step": 16418
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3615282350976825,
+      "learning_rate": 6.733928505924369e-06,
+      "loss": 0.4854,
+      "step": 16419
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.018095595547753,
+      "learning_rate": 6.731081275338369e-06,
+      "loss": 1.4062,
+      "step": 16420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3158315063941934,
+      "learning_rate": 6.728234553168939e-06,
+      "loss": 0.8223,
+      "step": 16421
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.17022016313738,
+      "learning_rate": 6.725388339495295e-06,
+      "loss": 0.8496,
+      "step": 16422
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8732967906093785,
+      "learning_rate": 6.722542634396661e-06,
+      "loss": 1.1699,
+      "step": 16423
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6110363064052344,
+      "learning_rate": 6.719697437952224e-06,
+      "loss": 1.0938,
+      "step": 16424
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9656614266857737,
+      "learning_rate": 6.7168527502411595e-06,
+      "loss": 1.4668,
+      "step": 16425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8939452848053842,
+      "learning_rate": 6.714008571342645e-06,
+      "loss": 0.918,
+      "step": 16426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.335963628616411,
+      "learning_rate": 6.711164901335823e-06,
+      "loss": 1.0039,
+      "step": 16427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.318319460915879,
+      "learning_rate": 6.7083217402998415e-06,
+      "loss": 1.1221,
+      "step": 16428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.274054358001547,
+      "learning_rate": 6.705479088313821e-06,
+      "loss": 0.9097,
+      "step": 16429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.705079746523174,
+      "learning_rate": 6.70263694545687e-06,
+      "loss": 1.2793,
+      "step": 16430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.032994304036882,
+      "learning_rate": 6.699795311808077e-06,
+      "loss": 1.2773,
+      "step": 16431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.571701803727158,
+      "learning_rate": 6.696954187446536e-06,
+      "loss": 1.0195,
+      "step": 16432
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9656905622239584,
+      "learning_rate": 6.694113572451311e-06,
+      "loss": 0.7789,
+      "step": 16433
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.913001800662269,
+      "learning_rate": 6.691273466901443e-06,
+      "loss": 0.4697,
+      "step": 16434
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4369899509553026,
+      "learning_rate": 6.688433870875982e-06,
+      "loss": 0.7344,
+      "step": 16435
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0955610271533898,
+      "learning_rate": 6.685594784453958e-06,
+      "loss": 0.8428,
+      "step": 16436
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3522878719981084,
+      "learning_rate": 6.682756207714377e-06,
+      "loss": 0.6475,
+      "step": 16437
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.980843576682598,
+      "learning_rate": 6.6799181407362355e-06,
+      "loss": 1.0068,
+      "step": 16438
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4343029476760805,
+      "learning_rate": 6.677080583598502e-06,
+      "loss": 0.7153,
+      "step": 16439
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1175142274729226,
+      "learning_rate": 6.674243536380168e-06,
+      "loss": 0.6694,
+      "step": 16440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.194669456328096,
+      "learning_rate": 6.671406999160174e-06,
+      "loss": 1.6221,
+      "step": 16441
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.413740513139411,
+      "learning_rate": 6.6685709720174626e-06,
+      "loss": 1.0017,
+      "step": 16442
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6463465152101278,
+      "learning_rate": 6.66573545503095e-06,
+      "loss": 0.2867,
+      "step": 16443
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.9692046012911835,
+      "learning_rate": 6.662900448279557e-06,
+      "loss": 1.3047,
+      "step": 16444
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2706906218408456,
+      "learning_rate": 6.660065951842184e-06,
+      "loss": 1.1514,
+      "step": 16445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7395165632303855,
+      "learning_rate": 6.657231965797714e-06,
+      "loss": 0.8516,
+      "step": 16446
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.8922531346823375,
+      "learning_rate": 6.654398490225009e-06,
+      "loss": 0.9189,
+      "step": 16447
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5076045725535074,
+      "learning_rate": 6.651565525202915e-06,
+      "loss": 1.3633,
+      "step": 16448
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.397679316691719,
+      "learning_rate": 6.648733070810295e-06,
+      "loss": 0.6875,
+      "step": 16449
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.153874627068908,
+      "learning_rate": 6.645901127125959e-06,
+      "loss": 0.835,
+      "step": 16450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3941721841523935,
+      "learning_rate": 6.643069694228726e-06,
+      "loss": 1.6055,
+      "step": 16451
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8381112335213947,
+      "learning_rate": 6.64023877219738e-06,
+      "loss": 1.2988,
+      "step": 16452
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8653246153727057,
+      "learning_rate": 6.637408361110717e-06,
+      "loss": 1.3135,
+      "step": 16453
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5609822302666583,
+      "learning_rate": 6.634578461047508e-06,
+      "loss": 0.8945,
+      "step": 16454
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4969376882005094,
+      "learning_rate": 6.631749072086507e-06,
+      "loss": 0.8662,
+      "step": 16455
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.7207808862836425,
+      "learning_rate": 6.6289201943064425e-06,
+      "loss": 1.2266,
+      "step": 16456
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6151205982532053,
+      "learning_rate": 6.62609182778606e-06,
+      "loss": 1.1934,
+      "step": 16457
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7550623737950493,
+      "learning_rate": 6.62326397260406e-06,
+      "loss": 1.6309,
+      "step": 16458
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7407800437595047,
+      "learning_rate": 6.620436628839144e-06,
+      "loss": 0.8223,
+      "step": 16459
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.560187925611057,
+      "learning_rate": 6.617609796569985e-06,
+      "loss": 1.8945,
+      "step": 16460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.136834281418454,
+      "learning_rate": 6.614783475875269e-06,
+      "loss": 0.8408,
+      "step": 16461
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.8453789709822,
+      "learning_rate": 6.611957666833638e-06,
+      "loss": 1.2139,
+      "step": 16462
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.241728488252292,
+      "learning_rate": 6.609132369523746e-06,
+      "loss": 1.1729,
+      "step": 16463
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.983394897148703,
+      "learning_rate": 6.606307584024216e-06,
+      "loss": 0.9111,
+      "step": 16464
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8538007125039893,
+      "learning_rate": 6.603483310413647e-06,
+      "loss": 0.6475,
+      "step": 16465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9709077723211443,
+      "learning_rate": 6.600659548770657e-06,
+      "loss": 1.0356,
+      "step": 16466
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.099946829713093,
+      "learning_rate": 6.597836299173821e-06,
+      "loss": 0.8994,
+      "step": 16467
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.815627245632212,
+      "learning_rate": 6.59501356170171e-06,
+      "loss": 0.7725,
+      "step": 16468
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1166305654055306,
+      "learning_rate": 6.592191336432871e-06,
+      "loss": 0.8882,
+      "step": 16469
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1470271249593655,
+      "learning_rate": 6.58936962344586e-06,
+      "loss": 1.2891,
+      "step": 16470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.298309087142178,
+      "learning_rate": 6.586548422819192e-06,
+      "loss": 1.0153,
+      "step": 16471
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.866999002544739,
+      "learning_rate": 6.5837277346313896e-06,
+      "loss": 0.8682,
+      "step": 16472
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.225123211310213,
+      "learning_rate": 6.5809075589609416e-06,
+      "loss": 1.1445,
+      "step": 16473
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7732908885699574,
+      "learning_rate": 6.578087895886345e-06,
+      "loss": 0.8159,
+      "step": 16474
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.3583455670173645,
+      "learning_rate": 6.575268745486063e-06,
+      "loss": 1.207,
+      "step": 16475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2292477126555235,
+      "learning_rate": 6.572450107838551e-06,
+      "loss": 0.7842,
+      "step": 16476
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.686848831560542,
+      "learning_rate": 6.5696319830222425e-06,
+      "loss": 0.9858,
+      "step": 16477
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.45164036660802,
+      "learning_rate": 6.566814371115582e-06,
+      "loss": 1.373,
+      "step": 16478
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.727363420385374,
+      "learning_rate": 6.563997272196973e-06,
+      "loss": 0.7734,
+      "step": 16479
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.712419184607676,
+      "learning_rate": 6.5611806863448114e-06,
+      "loss": 1.0586,
+      "step": 16480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3622346436365462,
+      "learning_rate": 6.5583646136374796e-06,
+      "loss": 1.2607,
+      "step": 16481
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8228960118587985,
+      "learning_rate": 6.555549054153354e-06,
+      "loss": 1.3047,
+      "step": 16482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9782791824878374,
+      "learning_rate": 6.552734007970798e-06,
+      "loss": 0.6201,
+      "step": 16483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5154416983029115,
+      "learning_rate": 6.549919475168142e-06,
+      "loss": 1.3779,
+      "step": 16484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5764433954122072,
+      "learning_rate": 6.547105455823718e-06,
+      "loss": 1.0684,
+      "step": 16485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.068093306479627,
+      "learning_rate": 6.544291950015827e-06,
+      "loss": 1.0059,
+      "step": 16486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.184375235458102,
+      "learning_rate": 6.541478957822789e-06,
+      "loss": 1.2344,
+      "step": 16487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.424071073075577,
+      "learning_rate": 6.538666479322872e-06,
+      "loss": 1.0342,
+      "step": 16488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.128432577099203,
+      "learning_rate": 6.535854514594356e-06,
+      "loss": 1.1504,
+      "step": 16489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.321226922384241,
+      "learning_rate": 6.533043063715474e-06,
+      "loss": 1.248,
+      "step": 16490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.42512462893051,
+      "learning_rate": 6.530232126764502e-06,
+      "loss": 1.1602,
+      "step": 16491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.69177341033004,
+      "learning_rate": 6.52742170381965e-06,
+      "loss": 1.0283,
+      "step": 16492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4900544413170596,
+      "learning_rate": 6.524611794959135e-06,
+      "loss": 1.584,
+      "step": 16493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5665311258771575,
+      "learning_rate": 6.52180240026114e-06,
+      "loss": 1.4766,
+      "step": 16494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4566108060006058,
+      "learning_rate": 6.518993519803873e-06,
+      "loss": 1.2793,
+      "step": 16495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1742223312879747,
+      "learning_rate": 6.516185153665491e-06,
+      "loss": 1.0444,
+      "step": 16496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.443988130005755,
+      "learning_rate": 6.513377301924151e-06,
+      "loss": 1.1152,
+      "step": 16497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5986361251256804,
+      "learning_rate": 6.5105699646579895e-06,
+      "loss": 1.4316,
+      "step": 16498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.814783302380984,
+      "learning_rate": 6.507763141945136e-06,
+      "loss": 0.8293,
+      "step": 16499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.016247936995837,
+      "learning_rate": 6.504956833863715e-06,
+      "loss": 0.9092,
+      "step": 16500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.090191874585207,
+      "learning_rate": 6.502151040491817e-06,
+      "loss": 1.2559,
+      "step": 16501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0585491620512775,
+      "learning_rate": 6.499345761907521e-06,
+      "loss": 1.0508,
+      "step": 16502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.154017080085213,
+      "learning_rate": 6.496540998188899e-06,
+      "loss": 0.9658,
+      "step": 16503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2482212774509613,
+      "learning_rate": 6.4937367494140115e-06,
+      "loss": 0.8447,
+      "step": 16504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.062930329898241,
+      "learning_rate": 6.490933015660899e-06,
+      "loss": 1.2705,
+      "step": 16505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.651219050266763,
+      "learning_rate": 6.488129797007584e-06,
+      "loss": 0.9443,
+      "step": 16506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1541858689456412,
+      "learning_rate": 6.485327093532073e-06,
+      "loss": 0.8481,
+      "step": 16507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.22657407651687,
+      "learning_rate": 6.482524905312373e-06,
+      "loss": 1.5938,
+      "step": 16508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1647913838017003,
+      "learning_rate": 6.479723232426474e-06,
+      "loss": 1.2148,
+      "step": 16509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6546897344821194,
+      "learning_rate": 6.476922074952335e-06,
+      "loss": 0.7856,
+      "step": 16510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7349644488825833,
+      "learning_rate": 6.474121432967908e-06,
+      "loss": 0.8013,
+      "step": 16511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.191879259991476,
+      "learning_rate": 6.471321306551148e-06,
+      "loss": 1.1426,
+      "step": 16512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.906675574204448,
+      "learning_rate": 6.468521695779975e-06,
+      "loss": 1.2461,
+      "step": 16513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6899009259163873,
+      "learning_rate": 6.465722600732296e-06,
+      "loss": 1.2188,
+      "step": 16514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.988311976688995,
+      "learning_rate": 6.462924021486006e-06,
+      "loss": 1.291,
+      "step": 16515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.162350873595578,
+      "learning_rate": 6.4601259581189985e-06,
+      "loss": 1.2959,
+      "step": 16516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5488529131038415,
+      "learning_rate": 6.4573284107091406e-06,
+      "loss": 1.2617,
+      "step": 16517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3737476460152394,
+      "learning_rate": 6.454531379334275e-06,
+      "loss": 1.0127,
+      "step": 16518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8109553101711295,
+      "learning_rate": 6.451734864072259e-06,
+      "loss": 1.0488,
+      "step": 16519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.361491950195098,
+      "learning_rate": 6.4489388650009055e-06,
+      "loss": 0.8936,
+      "step": 16520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.683471143828594,
+      "learning_rate": 6.446143382198036e-06,
+      "loss": 1.3086,
+      "step": 16521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6899589999169375,
+      "learning_rate": 6.443348415741446e-06,
+      "loss": 0.8691,
+      "step": 16522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.033086476044986,
+      "learning_rate": 6.440553965708912e-06,
+      "loss": 0.9365,
+      "step": 16523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.777654075928547,
+      "learning_rate": 6.4377600321782e-06,
+      "loss": 1.5039,
+      "step": 16524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.294175288253846,
+      "learning_rate": 6.434966615227075e-06,
+      "loss": 1.0566,
+      "step": 16525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2458654660816375,
+      "learning_rate": 6.432173714933274e-06,
+      "loss": 0.7344,
+      "step": 16526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4102202519801836,
+      "learning_rate": 6.429381331374512e-06,
+      "loss": 1.2109,
+      "step": 16527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4566672439091426,
+      "learning_rate": 6.426589464628513e-06,
+      "loss": 0.5415,
+      "step": 16528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.302981722350312,
+      "learning_rate": 6.423798114772961e-06,
+      "loss": 0.5835,
+      "step": 16529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.331713170734641,
+      "learning_rate": 6.421007281885552e-06,
+      "loss": 1.1367,
+      "step": 16530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.343023664824157,
+      "learning_rate": 6.41821696604395e-06,
+      "loss": 1.0977,
+      "step": 16531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.574554404983112,
+      "learning_rate": 6.415427167325794e-06,
+      "loss": 1.2285,
+      "step": 16532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.651431059422841,
+      "learning_rate": 6.412637885808742e-06,
+      "loss": 1.248,
+      "step": 16533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.614387367645587,
+      "learning_rate": 6.409849121570413e-06,
+      "loss": 0.8022,
+      "step": 16534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.859314989676295,
+      "learning_rate": 6.407060874688414e-06,
+      "loss": 0.8423,
+      "step": 16535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.131081119330842,
+      "learning_rate": 6.404273145240333e-06,
+      "loss": 0.9854,
+      "step": 16536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1809414227230786,
+      "learning_rate": 6.4014859333037606e-06,
+      "loss": 0.5261,
+      "step": 16537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.480524990092377,
+      "learning_rate": 6.398699238956271e-06,
+      "loss": 1.2715,
+      "step": 16538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8754852302842275,
+      "learning_rate": 6.39591306227541e-06,
+      "loss": 0.856,
+      "step": 16539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.881735653745169,
+      "learning_rate": 6.393127403338714e-06,
+      "loss": 0.7734,
+      "step": 16540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.253063519207537,
+      "learning_rate": 6.3903422622237e-06,
+      "loss": 0.9951,
+      "step": 16541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.929457998823352,
+      "learning_rate": 6.387557639007891e-06,
+      "loss": 0.7988,
+      "step": 16542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.009535588184963,
+      "learning_rate": 6.3847735337687754e-06,
+      "loss": 1.1934,
+      "step": 16543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.963940807936062,
+      "learning_rate": 6.381989946583836e-06,
+      "loss": 0.5107,
+      "step": 16544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8263897709658967,
+      "learning_rate": 6.37920687753053e-06,
+      "loss": 0.9189,
+      "step": 16545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7146668918623313,
+      "learning_rate": 6.376424326686314e-06,
+      "loss": 0.8621,
+      "step": 16546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.97085455414798,
+      "learning_rate": 6.373642294128637e-06,
+      "loss": 1.2988,
+      "step": 16547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.788986628617494,
+      "learning_rate": 6.370860779934912e-06,
+      "loss": 1.7285,
+      "step": 16548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9029630870877348,
+      "learning_rate": 6.368079784182545e-06,
+      "loss": 0.5264,
+      "step": 16549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8207572426453527,
+      "learning_rate": 6.365299306948927e-06,
+      "loss": 1.0918,
+      "step": 16550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.6552157693551575,
+      "learning_rate": 6.362519348311452e-06,
+      "loss": 1.3008,
+      "step": 16551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0824937145784004,
+      "learning_rate": 6.359739908347473e-06,
+      "loss": 0.8647,
+      "step": 16552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.945758212756473,
+      "learning_rate": 6.356960987134339e-06,
+      "loss": 0.8052,
+      "step": 16553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.083579952958192,
+      "learning_rate": 6.354182584749397e-06,
+      "loss": 1.4688,
+      "step": 16554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7328603505716005,
+      "learning_rate": 6.351404701269956e-06,
+      "loss": 0.7529,
+      "step": 16555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.179596054729406,
+      "learning_rate": 6.348627336773338e-06,
+      "loss": 1.2388,
+      "step": 16556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.876231287715666,
+      "learning_rate": 6.3458504913368275e-06,
+      "loss": 1.2305,
+      "step": 16557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4112591808850294,
+      "learning_rate": 6.343074165037696e-06,
+      "loss": 0.9222,
+      "step": 16558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.3141633066532945,
+      "learning_rate": 6.3402983579532254e-06,
+      "loss": 1.2686,
+      "step": 16559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6452185854866708,
+      "learning_rate": 6.337523070160653e-06,
+      "loss": 1.1868,
+      "step": 16560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9043642941791514,
+      "learning_rate": 6.334748301737217e-06,
+      "loss": 0.8457,
+      "step": 16561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.684989291191919,
+      "learning_rate": 6.331974052760128e-06,
+      "loss": 1.6582,
+      "step": 16562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0444666024799356,
+      "learning_rate": 6.3292003233066114e-06,
+      "loss": 1.1602,
+      "step": 16563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1830506988733265,
+      "learning_rate": 6.326427113453839e-06,
+      "loss": 0.5052,
+      "step": 16564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.211629063871454,
+      "learning_rate": 6.323654423279005e-06,
+      "loss": 1.0156,
+      "step": 16565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9427249352067064,
+      "learning_rate": 6.320882252859267e-06,
+      "loss": 0.9274,
+      "step": 16566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9857050143949917,
+      "learning_rate": 6.318110602271765e-06,
+      "loss": 0.3586,
+      "step": 16567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7638348319061796,
+      "learning_rate": 6.315339471593646e-06,
+      "loss": 0.8301,
+      "step": 16568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.169873859413349,
+      "learning_rate": 6.312568860902021e-06,
+      "loss": 0.6104,
+      "step": 16569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2366491606291627,
+      "learning_rate": 6.309798770273992e-06,
+      "loss": 0.9482,
+      "step": 16570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3539350840223396,
+      "learning_rate": 6.3070291997866595e-06,
+      "loss": 1.2402,
+      "step": 16571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.304898837469979,
+      "learning_rate": 6.304260149517097e-06,
+      "loss": 0.7939,
+      "step": 16572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.238391730630886,
+      "learning_rate": 6.301491619542354e-06,
+      "loss": 0.6071,
+      "step": 16573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.242065877558469,
+      "learning_rate": 6.298723609939497e-06,
+      "loss": 1.1289,
+      "step": 16574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.128223442679987,
+      "learning_rate": 6.29595612078554e-06,
+      "loss": 0.7842,
+      "step": 16575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.062196651460979,
+      "learning_rate": 6.2931891521575166e-06,
+      "loss": 0.7427,
+      "step": 16576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.1771364676232485,
+      "learning_rate": 6.290422704132424e-06,
+      "loss": 0.7402,
+      "step": 16577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.976761801202109,
+      "learning_rate": 6.287656776787251e-06,
+      "loss": 1.0977,
+      "step": 16578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5766305259544846,
+      "learning_rate": 6.284891370198964e-06,
+      "loss": 0.7437,
+      "step": 16579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.080675051325665,
+      "learning_rate": 6.282126484444539e-06,
+      "loss": 1.2422,
+      "step": 16580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5687400574388066,
+      "learning_rate": 6.279362119600915e-06,
+      "loss": 0.4453,
+      "step": 16581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.66880116473566,
+      "learning_rate": 6.276598275745021e-06,
+      "loss": 1.2832,
+      "step": 16582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.09657552431058,
+      "learning_rate": 6.2738349529537664e-06,
+      "loss": 0.5112,
+      "step": 16583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.746422691447202,
+      "learning_rate": 6.271072151304064e-06,
+      "loss": 1.4805,
+      "step": 16584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.148055385941097,
+      "learning_rate": 6.268309870872804e-06,
+      "loss": 0.8506,
+      "step": 16585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8151201186352552,
+      "learning_rate": 6.265548111736858e-06,
+      "loss": 0.9272,
+      "step": 16586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.354937510242454,
+      "learning_rate": 6.262786873973078e-06,
+      "loss": 0.8291,
+      "step": 16587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7372717635930255,
+      "learning_rate": 6.260026157658306e-06,
+      "loss": 1.3281,
+      "step": 16588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2613431611504353,
+      "learning_rate": 6.257265962869382e-06,
+      "loss": 1.2012,
+      "step": 16589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4991493553780577,
+      "learning_rate": 6.254506289683118e-06,
+      "loss": 0.9248,
+      "step": 16590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.063127205188988,
+      "learning_rate": 6.2517471381763035e-06,
+      "loss": 0.9082,
+      "step": 16591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9564138610556925,
+      "learning_rate": 6.248988508425735e-06,
+      "loss": 1.1035,
+      "step": 16592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.903983432091985,
+      "learning_rate": 6.246230400508191e-06,
+      "loss": 0.98,
+      "step": 16593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2116880187685837,
+      "learning_rate": 6.24347281450042e-06,
+      "loss": 0.9473,
+      "step": 16594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.862740720883298,
+      "learning_rate": 6.240715750479162e-06,
+      "loss": 1.6357,
+      "step": 16595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.582152671094333,
+      "learning_rate": 6.2379592085211435e-06,
+      "loss": 1.4883,
+      "step": 16596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3785943565408685,
+      "learning_rate": 6.23520318870309e-06,
+      "loss": 1.3955,
+      "step": 16597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1912697054232093,
+      "learning_rate": 6.232447691101689e-06,
+      "loss": 1.0303,
+      "step": 16598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2088396407751447,
+      "learning_rate": 6.229692715793631e-06,
+      "loss": 1.2402,
+      "step": 16599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6935140165619655,
+      "learning_rate": 6.2269382628555775e-06,
+      "loss": 1.0156,
+      "step": 16600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.671875071302751,
+      "learning_rate": 6.224184332364186e-06,
+      "loss": 0.8672,
+      "step": 16601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.15548847406536,
+      "learning_rate": 6.22143092439611e-06,
+      "loss": 1.0391,
+      "step": 16602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3579428665189446,
+      "learning_rate": 6.218678039027967e-06,
+      "loss": 1.627,
+      "step": 16603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4906722577636065,
+      "learning_rate": 6.215925676336367e-06,
+      "loss": 0.939,
+      "step": 16604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5338099215422334,
+      "learning_rate": 6.2131738363979035e-06,
+      "loss": 0.9199,
+      "step": 16605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7093635849233744,
+      "learning_rate": 6.210422519289169e-06,
+      "loss": 0.9971,
+      "step": 16606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.82675753983859,
+      "learning_rate": 6.207671725086727e-06,
+      "loss": 0.7207,
+      "step": 16607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.483775408008505,
+      "learning_rate": 6.2049214538671325e-06,
+      "loss": 1.124,
+      "step": 16608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9104076994418997,
+      "learning_rate": 6.202171705706913e-06,
+      "loss": 0.415,
+      "step": 16609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.515566087344677,
+      "learning_rate": 6.199422480682604e-06,
+      "loss": 1.3789,
+      "step": 16610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1560525112657736,
+      "learning_rate": 6.196673778870718e-06,
+      "loss": 0.958,
+      "step": 16611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.379680636533598,
+      "learning_rate": 6.19392560034775e-06,
+      "loss": 0.8584,
+      "step": 16612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4709323942537678,
+      "learning_rate": 6.191177945190166e-06,
+      "loss": 1.1768,
+      "step": 16613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6717215153741023,
+      "learning_rate": 6.1884308134744526e-06,
+      "loss": 0.6366,
+      "step": 16614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.205065032727678,
+      "learning_rate": 6.185684205277048e-06,
+      "loss": 0.9785,
+      "step": 16615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.106658362212426,
+      "learning_rate": 6.182938120674395e-06,
+      "loss": 1.5469,
+      "step": 16616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6404701503517667,
+      "learning_rate": 6.180192559742906e-06,
+      "loss": 1.166,
+      "step": 16617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5998852976433797,
+      "learning_rate": 6.177447522559004e-06,
+      "loss": 1.5801,
+      "step": 16618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1072344055793537,
+      "learning_rate": 6.174703009199073e-06,
+      "loss": 1.3926,
+      "step": 16619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9001154136739622,
+      "learning_rate": 6.17195901973949e-06,
+      "loss": 0.5752,
+      "step": 16620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5704006459646904,
+      "learning_rate": 6.169215554256627e-06,
+      "loss": 1.0186,
+      "step": 16621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8528110226934364,
+      "learning_rate": 6.166472612826824e-06,
+      "loss": 1.2031,
+      "step": 16622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.270550577371344,
+      "learning_rate": 6.163730195526424e-06,
+      "loss": 1.2559,
+      "step": 16623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.384077409798716,
+      "learning_rate": 6.160988302431747e-06,
+      "loss": 0.873,
+      "step": 16624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.332581932697714,
+      "learning_rate": 6.158246933619097e-06,
+      "loss": 1.502,
+      "step": 16625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8476987574983816,
+      "learning_rate": 6.155506089164756e-06,
+      "loss": 0.769,
+      "step": 16626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.966242874142296,
+      "learning_rate": 6.152765769145019e-06,
+      "loss": 0.6235,
+      "step": 16627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9142909322395054,
+      "learning_rate": 6.150025973636134e-06,
+      "loss": 1.459,
+      "step": 16628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1422287996124307,
+      "learning_rate": 6.147286702714347e-06,
+      "loss": 0.0875,
+      "step": 16629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.076198780682045,
+      "learning_rate": 6.144547956455895e-06,
+      "loss": 0.6655,
+      "step": 16630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.195290904637738,
+      "learning_rate": 6.141809734937007e-06,
+      "loss": 0.6501,
+      "step": 16631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.832195997828332,
+      "learning_rate": 6.139072038233876e-06,
+      "loss": 1.2676,
+      "step": 16632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1293190673071685,
+      "learning_rate": 6.136334866422691e-06,
+      "loss": 0.9629,
+      "step": 16633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.110487806872853,
+      "learning_rate": 6.13359821957962e-06,
+      "loss": 0.9434,
+      "step": 16634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6858953688115044,
+      "learning_rate": 6.130862097780837e-06,
+      "loss": 0.9736,
+      "step": 16635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.4505314284619475,
+      "learning_rate": 6.128126501102479e-06,
+      "loss": 0.9717,
+      "step": 16636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.287688065420143,
+      "learning_rate": 6.12539142962068e-06,
+      "loss": 1.3545,
+      "step": 16637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.113556059563821,
+      "learning_rate": 6.122656883411546e-06,
+      "loss": 0.957,
+      "step": 16638
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.472331768518681,
+      "learning_rate": 6.119922862551186e-06,
+      "loss": 1.0371,
+      "step": 16639
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2302155954787177,
+      "learning_rate": 6.117189367115694e-06,
+      "loss": 1.4971,
+      "step": 16640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.634854309030909,
+      "learning_rate": 6.114456397181134e-06,
+      "loss": 0.9131,
+      "step": 16641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7746233916542864,
+      "learning_rate": 6.111723952823564e-06,
+      "loss": 1.2988,
+      "step": 16642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.547170142174412,
+      "learning_rate": 6.108992034119021e-06,
+      "loss": 1.2344,
+      "step": 16643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9633356757007003,
+      "learning_rate": 6.106260641143546e-06,
+      "loss": 0.9126,
+      "step": 16644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0925196503467465,
+      "learning_rate": 6.103529773973143e-06,
+      "loss": 0.5596,
+      "step": 16645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.665228803735887,
+      "learning_rate": 6.100799432683818e-06,
+      "loss": 0.9644,
+      "step": 16646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.489373606595955,
+      "learning_rate": 6.098069617351534e-06,
+      "loss": 0.5636,
+      "step": 16647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1977501410617815,
+      "learning_rate": 6.095340328052293e-06,
+      "loss": 0.7891,
+      "step": 16648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9858096204648823,
+      "learning_rate": 6.092611564862033e-06,
+      "loss": 1.2891,
+      "step": 16649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9976606785053295,
+      "learning_rate": 6.089883327856699e-06,
+      "loss": 1.2773,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.123619542181615,
+      "learning_rate": 6.087155617112203e-06,
+      "loss": 1.0684,
+      "step": 16651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.684777766288657,
+      "learning_rate": 6.084428432704475e-06,
+      "loss": 1.0117,
+      "step": 16652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.688314574191491,
+      "learning_rate": 6.081701774709406e-06,
+      "loss": 1.584,
+      "step": 16653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.287077371713804,
+      "learning_rate": 6.0789756432028724e-06,
+      "loss": 1.0098,
+      "step": 16654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6739087312083094,
+      "learning_rate": 6.076250038260736e-06,
+      "loss": 0.9473,
+      "step": 16655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5839521637638825,
+      "learning_rate": 6.073524959958865e-06,
+      "loss": 1.7695,
+      "step": 16656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0495062070530605,
+      "learning_rate": 6.070800408373081e-06,
+      "loss": 0.6602,
+      "step": 16657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.427873246562735,
+      "learning_rate": 6.068076383579224e-06,
+      "loss": 0.6731,
+      "step": 16658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0990978664720097,
+      "learning_rate": 6.065352885653094e-06,
+      "loss": 0.9414,
+      "step": 16659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.759783965501701,
+      "learning_rate": 6.062629914670478e-06,
+      "loss": 1.4453,
+      "step": 16660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1970569726606355,
+      "learning_rate": 6.059907470707171e-06,
+      "loss": 1.0527,
+      "step": 16661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.115226109413016,
+      "learning_rate": 6.057185553838926e-06,
+      "loss": 1.8008,
+      "step": 16662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.912678980751169,
+      "learning_rate": 6.054464164141496e-06,
+      "loss": 1.2324,
+      "step": 16663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.061487986206994,
+      "learning_rate": 6.05174330169061e-06,
+      "loss": 1.2969,
+      "step": 16664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.936891695711033,
+      "learning_rate": 6.049022966561998e-06,
+      "loss": 1.1973,
+      "step": 16665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.497480076127106,
+      "learning_rate": 6.046303158831357e-06,
+      "loss": 0.4834,
+      "step": 16666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2589725469022293,
+      "learning_rate": 6.043583878574391e-06,
+      "loss": 1.1855,
+      "step": 16667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7965629019903733,
+      "learning_rate": 6.04086512586676e-06,
+      "loss": 1.0488,
+      "step": 16668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5583222922253595,
+      "learning_rate": 6.038146900784142e-06,
+      "loss": 0.666,
+      "step": 16669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.824030366425294,
+      "learning_rate": 6.035429203402176e-06,
+      "loss": 1.21,
+      "step": 16670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.734278427047215,
+      "learning_rate": 6.0327120337964945e-06,
+      "loss": 0.9463,
+      "step": 16671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.004720938693219,
+      "learning_rate": 6.029995392042709e-06,
+      "loss": 0.936,
+      "step": 16672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.172251058171915,
+      "learning_rate": 6.027279278216436e-06,
+      "loss": 1.0234,
+      "step": 16673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.815642789047887,
+      "learning_rate": 6.024563692393256e-06,
+      "loss": 1.2859,
+      "step": 16674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3649127971548913,
+      "learning_rate": 6.0218486346487355e-06,
+      "loss": 1.1719,
+      "step": 16675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6294997338719104,
+      "learning_rate": 6.019134105058452e-06,
+      "loss": 1.4082,
+      "step": 16676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2259424249652673,
+      "learning_rate": 6.016420103697931e-06,
+      "loss": 1.1523,
+      "step": 16677
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.97072665575722,
+      "learning_rate": 6.013706630642715e-06,
+      "loss": 1.0698,
+      "step": 16678
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.476909098097909,
+      "learning_rate": 6.010993685968317e-06,
+      "loss": 0.9678,
+      "step": 16679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.280838995369069,
+      "learning_rate": 6.008281269750233e-06,
+      "loss": 0.6179,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3185484744163043,
+      "learning_rate": 6.0055693820639415e-06,
+      "loss": 0.542,
+      "step": 16681
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.821994709497087,
+      "learning_rate": 6.002858022984931e-06,
+      "loss": 1.3457,
+      "step": 16682
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.462255066205753,
+      "learning_rate": 6.0001471925886485e-06,
+      "loss": 1.2004,
+      "step": 16683
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.32833638149473,
+      "learning_rate": 5.99743689095053e-06,
+      "loss": 1.0117,
+      "step": 16684
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3879156779381105,
+      "learning_rate": 5.994727118146004e-06,
+      "loss": 0.8296,
+      "step": 16685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.42576250306323,
+      "learning_rate": 5.992017874250486e-06,
+      "loss": 1.0957,
+      "step": 16686
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.589868377629885,
+      "learning_rate": 5.989309159339376e-06,
+      "loss": 1.7344,
+      "step": 16687
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6860983098630293,
+      "learning_rate": 5.986600973488055e-06,
+      "loss": 0.853,
+      "step": 16688
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.239340393835416,
+      "learning_rate": 5.983893316771879e-06,
+      "loss": 1.1812,
+      "step": 16689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4032192216714,
+      "learning_rate": 5.981186189266222e-06,
+      "loss": 1.2812,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.892409043955475,
+      "learning_rate": 5.978479591046407e-06,
+      "loss": 1.1445,
+      "step": 16691
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.452283491967541,
+      "learning_rate": 5.975773522187763e-06,
+      "loss": 1.4492,
+      "step": 16692
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.60286627281685,
+      "learning_rate": 5.97306798276559e-06,
+      "loss": 1.0947,
+      "step": 16693
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.1793955120256685,
+      "learning_rate": 5.9703629728551894e-06,
+      "loss": 1.3877,
+      "step": 16694
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.370936254793177,
+      "learning_rate": 5.967658492531847e-06,
+      "loss": 1.9121,
+      "step": 16695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6236077204015347,
+      "learning_rate": 5.964954541870821e-06,
+      "loss": 0.8184,
+      "step": 16696
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0047530033710075,
+      "learning_rate": 5.96225112094736e-06,
+      "loss": 1.1562,
+      "step": 16697
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.545882704568764,
+      "learning_rate": 5.959548229836692e-06,
+      "loss": 1.2422,
+      "step": 16698
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8087340679319523,
+      "learning_rate": 5.956845868614055e-06,
+      "loss": 0.4336,
+      "step": 16699
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4362918109153817,
+      "learning_rate": 5.954144037354645e-06,
+      "loss": 1.0957,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5013753625521382,
+      "learning_rate": 5.95144273613365e-06,
+      "loss": 0.511,
+      "step": 16701
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.653771264902482,
+      "learning_rate": 5.9487419650262444e-06,
+      "loss": 1.1035,
+      "step": 16702
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.915599532837093,
+      "learning_rate": 5.946041724107593e-06,
+      "loss": 0.7031,
+      "step": 16703
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3092746408256106,
+      "learning_rate": 5.943342013452852e-06,
+      "loss": 0.9658,
+      "step": 16704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4416375866960367,
+      "learning_rate": 5.940642833137144e-06,
+      "loss": 1.502,
+      "step": 16705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2263737731862685,
+      "learning_rate": 5.9379441832355855e-06,
+      "loss": 1.4062,
+      "step": 16706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6336718783594204,
+      "learning_rate": 5.935246063823274e-06,
+      "loss": 0.9072,
+      "step": 16707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.847473962697477,
+      "learning_rate": 5.932548474975311e-06,
+      "loss": 1.8125,
+      "step": 16708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.470050614027606,
+      "learning_rate": 5.929851416766763e-06,
+      "loss": 0.9062,
+      "step": 16709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1969264061069724,
+      "learning_rate": 5.927154889272677e-06,
+      "loss": 1.2627,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.148454311127187,
+      "learning_rate": 5.924458892568114e-06,
+      "loss": 1.3828,
+      "step": 16711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.234992063260439,
+      "learning_rate": 5.921763426728086e-06,
+      "loss": 1.1562,
+      "step": 16712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.126764555855215,
+      "learning_rate": 5.919068491827623e-06,
+      "loss": 1.2422,
+      "step": 16713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9739193532407833,
+      "learning_rate": 5.916374087941717e-06,
+      "loss": 1.3105,
+      "step": 16714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.340231755643032,
+      "learning_rate": 5.913680215145342e-06,
+      "loss": 2.4893,
+      "step": 16715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2917824687020576,
+      "learning_rate": 5.910986873513486e-06,
+      "loss": 1.2383,
+      "step": 16716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7714252747893915,
+      "learning_rate": 5.908294063121095e-06,
+      "loss": 0.7935,
+      "step": 16717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.294235406523487,
+      "learning_rate": 5.905601784043105e-06,
+      "loss": 1.2764,
+      "step": 16718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.996692425854101,
+      "learning_rate": 5.9029100363544395e-06,
+      "loss": 0.6531,
+      "step": 16719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.317872926151536,
+      "learning_rate": 5.900218820130019e-06,
+      "loss": 1.1152,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.383856934637966,
+      "learning_rate": 5.897528135444733e-06,
+      "loss": 1.3691,
+      "step": 16721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.667441275467718,
+      "learning_rate": 5.894837982373455e-06,
+      "loss": 1.1123,
+      "step": 16722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2741273442949645,
+      "learning_rate": 5.892148360991065e-06,
+      "loss": 1.4746,
+      "step": 16723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.657205237557987,
+      "learning_rate": 5.8894592713724025e-06,
+      "loss": 1.3926,
+      "step": 16724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.341713185600603,
+      "learning_rate": 5.8867707135923124e-06,
+      "loss": 0.9854,
+      "step": 16725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4929696728306907,
+      "learning_rate": 5.884082687725615e-06,
+      "loss": 1.0312,
+      "step": 16726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6640986784427914,
+      "learning_rate": 5.881395193847106e-06,
+      "loss": 1.0244,
+      "step": 16727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.933682221934076,
+      "learning_rate": 5.878708232031593e-06,
+      "loss": 1.5791,
+      "step": 16728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.066229592841284,
+      "learning_rate": 5.876021802353846e-06,
+      "loss": 0.6484,
+      "step": 16729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9887078157045424,
+      "learning_rate": 5.873335904888627e-06,
+      "loss": 0.9536,
+      "step": 16730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.079649516617883,
+      "learning_rate": 5.870650539710676e-06,
+      "loss": 0.6677,
+      "step": 16731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3253309673136737,
+      "learning_rate": 5.867965706894732e-06,
+      "loss": 0.7036,
+      "step": 16732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.487077215102711,
+      "learning_rate": 5.865281406515521e-06,
+      "loss": 1.2515,
+      "step": 16733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8818512041455528,
+      "learning_rate": 5.862597638647738e-06,
+      "loss": 1.0947,
+      "step": 16734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.968431636404127,
+      "learning_rate": 5.8599144033660735e-06,
+      "loss": 1.0073,
+      "step": 16735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7309111535542816,
+      "learning_rate": 5.8572317007451914e-06,
+      "loss": 0.8545,
+      "step": 16736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.635887939270495,
+      "learning_rate": 5.854549530859765e-06,
+      "loss": 0.8945,
+      "step": 16737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.966618143892119,
+      "learning_rate": 5.851867893784432e-06,
+      "loss": 0.8047,
+      "step": 16738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.154530371392662,
+      "learning_rate": 5.849186789593819e-06,
+      "loss": 0.8594,
+      "step": 16739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.862752784994821,
+      "learning_rate": 5.846506218362535e-06,
+      "loss": 0.7144,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.015350451156844,
+      "learning_rate": 5.843826180165188e-06,
+      "loss": 1.1289,
+      "step": 16741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.881719614678373,
+      "learning_rate": 5.841146675076364e-06,
+      "loss": 1.0215,
+      "step": 16742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9634017151479477,
+      "learning_rate": 5.8384677031706275e-06,
+      "loss": 1.1592,
+      "step": 16743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5379515666331267,
+      "learning_rate": 5.835789264522537e-06,
+      "loss": 0.5977,
+      "step": 16744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.752506799266593,
+      "learning_rate": 5.83311135920662e-06,
+      "loss": 0.835,
+      "step": 16745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.33271470270505,
+      "learning_rate": 5.83043398729742e-06,
+      "loss": 0.4712,
+      "step": 16746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3513670110710807,
+      "learning_rate": 5.8277571488694365e-06,
+      "loss": 0.8135,
+      "step": 16747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8425389320572605,
+      "learning_rate": 5.825080843997161e-06,
+      "loss": 0.7515,
+      "step": 16748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8992752550766383,
+      "learning_rate": 5.82240507275508e-06,
+      "loss": 0.9824,
+      "step": 16749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.396038153096654,
+      "learning_rate": 5.819729835217663e-06,
+      "loss": 1.2012,
+      "step": 16750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.859816673412201,
+      "learning_rate": 5.817055131459359e-06,
+      "loss": 1.1592,
+      "step": 16751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3308429633955665,
+      "learning_rate": 5.8143809615545995e-06,
+      "loss": 0.7144,
+      "step": 16752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3303361451543716,
+      "learning_rate": 5.8117073255778e-06,
+      "loss": 1.3887,
+      "step": 16753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.51128581437975,
+      "learning_rate": 5.809034223603382e-06,
+      "loss": 1.1152,
+      "step": 16754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.232338020432439,
+      "learning_rate": 5.80636165570573e-06,
+      "loss": 1.7889,
+      "step": 16755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.22637814950026,
+      "learning_rate": 5.803689621959219e-06,
+      "loss": 1.5762,
+      "step": 16756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4273632253817703,
+      "learning_rate": 5.8010181224382035e-06,
+      "loss": 1.4707,
+      "step": 16757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.472808702027902,
+      "learning_rate": 5.798347157217044e-06,
+      "loss": 0.8145,
+      "step": 16758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.230136719690787,
+      "learning_rate": 5.795676726370061e-06,
+      "loss": 1.0015,
+      "step": 16759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9642206273508562,
+      "learning_rate": 5.793006829971581e-06,
+      "loss": 1.458,
+      "step": 16760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.351525433265792,
+      "learning_rate": 5.790337468095905e-06,
+      "loss": 1.2305,
+      "step": 16761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.07922231592724,
+      "learning_rate": 5.787668640817307e-06,
+      "loss": 1.1865,
+      "step": 16762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.438142838440945,
+      "learning_rate": 5.785000348210079e-06,
+      "loss": 1.3145,
+      "step": 16763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.636179161811138,
+      "learning_rate": 5.782332590348466e-06,
+      "loss": 1.127,
+      "step": 16764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.574646238451766,
+      "learning_rate": 5.779665367306717e-06,
+      "loss": 1.3232,
+      "step": 16765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6648282720310836,
+      "learning_rate": 5.7769986791590456e-06,
+      "loss": 0.7505,
+      "step": 16766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.07664299228373,
+      "learning_rate": 5.7743325259796835e-06,
+      "loss": 0.7549,
+      "step": 16767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3095748546466965,
+      "learning_rate": 5.771666907842815e-06,
+      "loss": 0.7339,
+      "step": 16768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.577077700241482,
+      "learning_rate": 5.769001824822634e-06,
+      "loss": 1.6094,
+      "step": 16769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8795041171304634,
+      "learning_rate": 5.766337276993297e-06,
+      "loss": 0.6306,
+      "step": 16770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1223882824082922,
+      "learning_rate": 5.7636732644289705e-06,
+      "loss": 0.9033,
+      "step": 16771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5951261431660875,
+      "learning_rate": 5.761009787203783e-06,
+      "loss": 0.9551,
+      "step": 16772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2373211844413756,
+      "learning_rate": 5.758346845391865e-06,
+      "loss": 1.1074,
+      "step": 16773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.300189371191221,
+      "learning_rate": 5.7556844390673135e-06,
+      "loss": 0.6875,
+      "step": 16774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.61517155717935,
+      "learning_rate": 5.753022568304233e-06,
+      "loss": 0.6709,
+      "step": 16775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7353368617923333,
+      "learning_rate": 5.750361233176704e-06,
+      "loss": 1.0081,
+      "step": 16776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.587331620786959,
+      "learning_rate": 5.747700433758777e-06,
+      "loss": 1.3511,
+      "step": 16777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.086379857288166,
+      "learning_rate": 5.745040170124513e-06,
+      "loss": 1.2109,
+      "step": 16778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.390330607903378,
+      "learning_rate": 5.7423804423479385e-06,
+      "loss": 1.084,
+      "step": 16779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7159949187903987,
+      "learning_rate": 5.739721250503083e-06,
+      "loss": 1.543,
+      "step": 16780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.814427623282649,
+      "learning_rate": 5.737062594663944e-06,
+      "loss": 0.6812,
+      "step": 16781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1722495215809783,
+      "learning_rate": 5.734404474904511e-06,
+      "loss": 0.7012,
+      "step": 16782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5743731393087304,
+      "learning_rate": 5.731746891298753e-06,
+      "loss": 1.0332,
+      "step": 16783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8572079508241317,
+      "learning_rate": 5.7290898439206406e-06,
+      "loss": 0.2703,
+      "step": 16784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.15163791000798,
+      "learning_rate": 5.726433332844111e-06,
+      "loss": 0.9277,
+      "step": 16785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.460559149828239,
+      "learning_rate": 5.72377735814309e-06,
+      "loss": 0.6161,
+      "step": 16786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.719489249212214,
+      "learning_rate": 5.721121919891498e-06,
+      "loss": 0.8892,
+      "step": 16787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.308148387931957,
+      "learning_rate": 5.718467018163243e-06,
+      "loss": 0.3164,
+      "step": 16788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.841921984172857,
+      "learning_rate": 5.715812653032199e-06,
+      "loss": 0.8613,
+      "step": 16789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6027103340962356,
+      "learning_rate": 5.713158824572237e-06,
+      "loss": 0.9062,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8546208743281998,
+      "learning_rate": 5.710505532857211e-06,
+      "loss": 1.541,
+      "step": 16791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8723224942945507,
+      "learning_rate": 5.707852777960965e-06,
+      "loss": 1.2129,
+      "step": 16792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0381432576840064,
+      "learning_rate": 5.705200559957324e-06,
+      "loss": 0.8145,
+      "step": 16793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.13892612386508,
+      "learning_rate": 5.7025488789201e-06,
+      "loss": 0.8271,
+      "step": 16794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9857441422852893,
+      "learning_rate": 5.6998977349230735e-06,
+      "loss": 0.4524,
+      "step": 16795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.377994857615693,
+      "learning_rate": 5.697247128040037e-06,
+      "loss": 0.6836,
+      "step": 16796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.324898243800072,
+      "learning_rate": 5.694597058344764e-06,
+      "loss": 0.623,
+      "step": 16797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1192606291978504,
+      "learning_rate": 5.691947525910992e-06,
+      "loss": 1.2559,
+      "step": 16798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.340939058239431,
+      "learning_rate": 5.6892985308124616e-06,
+      "loss": 1.0986,
+      "step": 16799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5855444059237596,
+      "learning_rate": 5.686650073122887e-06,
+      "loss": 1.0791,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7047048817840205,
+      "learning_rate": 5.684002152915982e-06,
+      "loss": 0.9167,
+      "step": 16801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3017215630866525,
+      "learning_rate": 5.681354770265434e-06,
+      "loss": 0.9521,
+      "step": 16802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.36253783638622,
+      "learning_rate": 5.678707925244919e-06,
+      "loss": 0.6514,
+      "step": 16803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.111737750526116,
+      "learning_rate": 5.676061617928091e-06,
+      "loss": 1.1201,
+      "step": 16804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.504064631592936,
+      "learning_rate": 5.6734158483886e-06,
+      "loss": 0.9258,
+      "step": 16805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.786353684679876,
+      "learning_rate": 5.6707706167000856e-06,
+      "loss": 1.3965,
+      "step": 16806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0292296705967696,
+      "learning_rate": 5.668125922936154e-06,
+      "loss": 1.0176,
+      "step": 16807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8913744867455358,
+      "learning_rate": 5.665481767170405e-06,
+      "loss": 1.0322,
+      "step": 16808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9183175415690066,
+      "learning_rate": 5.662838149476432e-06,
+      "loss": 1.0503,
+      "step": 16809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0016006978281515,
+      "learning_rate": 5.660195069927801e-06,
+      "loss": 1.5391,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.060141862595524,
+      "learning_rate": 5.6575525285980675e-06,
+      "loss": 1.124,
+      "step": 16811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.402955040600639,
+      "learning_rate": 5.654910525560766e-06,
+      "loss": 1.5664,
+      "step": 16812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2075175331948764,
+      "learning_rate": 5.652269060889437e-06,
+      "loss": 0.873,
+      "step": 16813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.893563164936031,
+      "learning_rate": 5.649628134657575e-06,
+      "loss": 1.3667,
+      "step": 16814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9106399031821324,
+      "learning_rate": 5.646987746938693e-06,
+      "loss": 0.3477,
+      "step": 16815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.945341803205108,
+      "learning_rate": 5.644347897806263e-06,
+      "loss": 1.3945,
+      "step": 16816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.615347408627669,
+      "learning_rate": 5.641708587333744e-06,
+      "loss": 1.2617,
+      "step": 16817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.057152027911983,
+      "learning_rate": 5.639069815594603e-06,
+      "loss": 0.9417,
+      "step": 16818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3839298817736605,
+      "learning_rate": 5.636431582662266e-06,
+      "loss": 1.293,
+      "step": 16819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1271416001521812,
+      "learning_rate": 5.6337938886101534e-06,
+      "loss": 0.562,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6634018710482845,
+      "learning_rate": 5.631156733511666e-06,
+      "loss": 0.6484,
+      "step": 16821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4439132635892604,
+      "learning_rate": 5.628520117440211e-06,
+      "loss": 1.4995,
+      "step": 16822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.689708909565228,
+      "learning_rate": 5.625884040469154e-06,
+      "loss": 1.1094,
+      "step": 16823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8735902744585524,
+      "learning_rate": 5.623248502671852e-06,
+      "loss": 1.0911,
+      "step": 16824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1265652445621983,
+      "learning_rate": 5.6206135041216524e-06,
+      "loss": 1.0303,
+      "step": 16825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.565299958267107,
+      "learning_rate": 5.6179790448918995e-06,
+      "loss": 1.0537,
+      "step": 16826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9007613333694633,
+      "learning_rate": 5.615345125055896e-06,
+      "loss": 1.2002,
+      "step": 16827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.774671396264882,
+      "learning_rate": 5.6127117446869495e-06,
+      "loss": 0.606,
+      "step": 16828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6386280804170898,
+      "learning_rate": 5.610078903858334e-06,
+      "loss": 0.6309,
+      "step": 16829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8304482712243058,
+      "learning_rate": 5.607446602643334e-06,
+      "loss": 0.9949,
+      "step": 16830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0214563045303473,
+      "learning_rate": 5.604814841115205e-06,
+      "loss": 0.8735,
+      "step": 16831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.974625192158151,
+      "learning_rate": 5.602183619347179e-06,
+      "loss": 1.0879,
+      "step": 16832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.86577428223686,
+      "learning_rate": 5.599552937412481e-06,
+      "loss": 0.7831,
+      "step": 16833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.401094437444741,
+      "learning_rate": 5.596922795384326e-06,
+      "loss": 0.9785,
+      "step": 16834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1834075793777776,
+      "learning_rate": 5.59429319333592e-06,
+      "loss": 0.8503,
+      "step": 16835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.304585588913891,
+      "learning_rate": 5.5916641313404325e-06,
+      "loss": 1.1875,
+      "step": 16836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6473056254583995,
+      "learning_rate": 5.58903560947103e-06,
+      "loss": 1.0703,
+      "step": 16837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.500600218396566,
+      "learning_rate": 5.5864076278008606e-06,
+      "loss": 0.6278,
+      "step": 16838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.951560277014118,
+      "learning_rate": 5.583780186403068e-06,
+      "loss": 0.9971,
+      "step": 16839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.607346398142802,
+      "learning_rate": 5.581153285350768e-06,
+      "loss": 0.7495,
+      "step": 16840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0616896222172127,
+      "learning_rate": 5.5785269247170675e-06,
+      "loss": 0.5044,
+      "step": 16841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.423056939886875,
+      "learning_rate": 5.575901104575049e-06,
+      "loss": 0.7207,
+      "step": 16842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9956023027447665,
+      "learning_rate": 5.573275824997795e-06,
+      "loss": 0.7314,
+      "step": 16843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8831150850408584,
+      "learning_rate": 5.570651086058373e-06,
+      "loss": 1.5078,
+      "step": 16844
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.447773925653977,
+      "learning_rate": 5.568026887829824e-06,
+      "loss": 0.6797,
+      "step": 16845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.2760721393758,
+      "learning_rate": 5.565403230385166e-06,
+      "loss": 1.0762,
+      "step": 16846
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.613318929129757,
+      "learning_rate": 5.562780113797434e-06,
+      "loss": 0.8735,
+      "step": 16847
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.115848875276529,
+      "learning_rate": 5.560157538139615e-06,
+      "loss": 1.1748,
+      "step": 16848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9187317862873254,
+      "learning_rate": 5.557535503484701e-06,
+      "loss": 0.9453,
+      "step": 16849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.420145511996502,
+      "learning_rate": 5.554914009905651e-06,
+      "loss": 1.1816,
+      "step": 16850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.3392811335771615,
+      "learning_rate": 5.552293057475428e-06,
+      "loss": 1.2832,
+      "step": 16851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0451490973703734,
+      "learning_rate": 5.549672646266979e-06,
+      "loss": 0.9531,
+      "step": 16852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.636076988022005,
+      "learning_rate": 5.547052776353221e-06,
+      "loss": 2.0137,
+      "step": 16853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.897499637729374,
+      "learning_rate": 5.544433447807065e-06,
+      "loss": 0.8684,
+      "step": 16854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.520444090290516,
+      "learning_rate": 5.5418146607013995e-06,
+      "loss": 1.1338,
+      "step": 16855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5629549320469165,
+      "learning_rate": 5.539196415109119e-06,
+      "loss": 1.0352,
+      "step": 16856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.0920600023540965,
+      "learning_rate": 5.536578711103077e-06,
+      "loss": 1.0967,
+      "step": 16857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9781824642952652,
+      "learning_rate": 5.533961548756128e-06,
+      "loss": 1.1533,
+      "step": 16858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7250964721885818,
+      "learning_rate": 5.5313449281410964e-06,
+      "loss": 1.4277,
+      "step": 16859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.507397008208892,
+      "learning_rate": 5.528728849330816e-06,
+      "loss": 1.3613,
+      "step": 16860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.897030012509013,
+      "learning_rate": 5.526113312398079e-06,
+      "loss": 1.2959,
+      "step": 16861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6595230450478553,
+      "learning_rate": 5.523498317415688e-06,
+      "loss": 0.8086,
+      "step": 16862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.23848863210171,
+      "learning_rate": 5.520883864456402e-06,
+      "loss": 1.377,
+      "step": 16863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.813069535536549,
+      "learning_rate": 5.5182699535929975e-06,
+      "loss": 1.6055,
+      "step": 16864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0400828881347457,
+      "learning_rate": 5.515656584898204e-06,
+      "loss": 1.0723,
+      "step": 16865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.951462240618995,
+      "learning_rate": 5.513043758444761e-06,
+      "loss": 1.0469,
+      "step": 16866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8131143043766977,
+      "learning_rate": 5.510431474305369e-06,
+      "loss": 0.9282,
+      "step": 16867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.839745645443732,
+      "learning_rate": 5.507819732552741e-06,
+      "loss": 0.5664,
+      "step": 16868
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.405187601650215,
+      "learning_rate": 5.5052085332595535e-06,
+      "loss": 0.9902,
+      "step": 16869
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.194186536170404,
+      "learning_rate": 5.5025978764984695e-06,
+      "loss": 1.0007,
+      "step": 16870
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.906555276409046,
+      "learning_rate": 5.4999877623421566e-06,
+      "loss": 1.1455,
+      "step": 16871
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.712682823534613,
+      "learning_rate": 5.497378190863239e-06,
+      "loss": 0.9434,
+      "step": 16872
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.4965473088217145,
+      "learning_rate": 5.4947691621343556e-06,
+      "loss": 0.7776,
+      "step": 16873
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.858371746224031,
+      "learning_rate": 5.492160676228103e-06,
+      "loss": 1.2832,
+      "step": 16874
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7323909864118114,
+      "learning_rate": 5.48955273321708e-06,
+      "loss": 0.9619,
+      "step": 16875
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3252328407434546,
+      "learning_rate": 5.486945333173851e-06,
+      "loss": 0.8643,
+      "step": 16876
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6206772042626025,
+      "learning_rate": 5.484338476170999e-06,
+      "loss": 1.1113,
+      "step": 16877
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.619277347463802,
+      "learning_rate": 5.481732162281064e-06,
+      "loss": 0.6582,
+      "step": 16878
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.954659350545424,
+      "learning_rate": 5.479126391576566e-06,
+      "loss": 1.2676,
+      "step": 16879
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0146522244571776,
+      "learning_rate": 5.476521164130044e-06,
+      "loss": 0.8865,
+      "step": 16880
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.2909793982454625,
+      "learning_rate": 5.473916480013983e-06,
+      "loss": 0.9814,
+      "step": 16881
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.474436736221143,
+      "learning_rate": 5.47131233930088e-06,
+      "loss": 1.1865,
+      "step": 16882
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3886281568592302,
+      "learning_rate": 5.4687087420632086e-06,
+      "loss": 1.0469,
+      "step": 16883
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.22056392112694,
+      "learning_rate": 5.466105688373413e-06,
+      "loss": 1.2695,
+      "step": 16884
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.089665998938965,
+      "learning_rate": 5.4635031783039524e-06,
+      "loss": 0.6073,
+      "step": 16885
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.72387151537239,
+      "learning_rate": 5.460901211927244e-06,
+      "loss": 1.1719,
+      "step": 16886
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.293245675700798,
+      "learning_rate": 5.458299789315699e-06,
+      "loss": 1.4277,
+      "step": 16887
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3696639384585036,
+      "learning_rate": 5.4556989105417095e-06,
+      "loss": 0.7587,
+      "step": 16888
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.133153544381765,
+      "learning_rate": 5.4530985756776644e-06,
+      "loss": 1.5645,
+      "step": 16889
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.776922996186914,
+      "learning_rate": 5.450498784795935e-06,
+      "loss": 1.3306,
+      "step": 16890
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8441160452896865,
+      "learning_rate": 5.4478995379688645e-06,
+      "loss": 0.6592,
+      "step": 16891
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7265973895007627,
+      "learning_rate": 5.445300835268791e-06,
+      "loss": 1.2051,
+      "step": 16892
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.303525896384573,
+      "learning_rate": 5.442702676768027e-06,
+      "loss": 1.0,
+      "step": 16893
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7990523233214857,
+      "learning_rate": 5.440105062538895e-06,
+      "loss": 1.0732,
+      "step": 16894
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6256578620174458,
+      "learning_rate": 5.437507992653673e-06,
+      "loss": 0.6506,
+      "step": 16895
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.350227357133455,
+      "learning_rate": 5.434911467184639e-06,
+      "loss": 1.4902,
+      "step": 16896
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.279364796779126,
+      "learning_rate": 5.432315486204051e-06,
+      "loss": 1.1123,
+      "step": 16897
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6646163806650596,
+      "learning_rate": 5.4297200497841535e-06,
+      "loss": 0.7229,
+      "step": 16898
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.880119203669704,
+      "learning_rate": 5.42712515799719e-06,
+      "loss": 0.6387,
+      "step": 16899
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.155695968437433,
+      "learning_rate": 5.424530810915362e-06,
+      "loss": 1.0684,
+      "step": 16900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6862849660437598,
+      "learning_rate": 5.421937008610875e-06,
+      "loss": 1.1875,
+      "step": 16901
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.442351770032135,
+      "learning_rate": 5.419343751155903e-06,
+      "loss": 0.8032,
+      "step": 16902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.311423634833891,
+      "learning_rate": 5.416751038622628e-06,
+      "loss": 2.2695,
+      "step": 16903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.114601949475154,
+      "learning_rate": 5.4141588710832025e-06,
+      "loss": 0.7832,
+      "step": 16904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.995874423126371,
+      "learning_rate": 5.4115672486097525e-06,
+      "loss": 0.6865,
+      "step": 16905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.555944333742486,
+      "learning_rate": 5.408976171274421e-06,
+      "loss": 1.002,
+      "step": 16906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.103599754462443,
+      "learning_rate": 5.4063856391493e-06,
+      "loss": 0.9219,
+      "step": 16907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9631173758095892,
+      "learning_rate": 5.403795652306495e-06,
+      "loss": 1.1084,
+      "step": 16908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.804420916169142,
+      "learning_rate": 5.40120621081808e-06,
+      "loss": 0.6067,
+      "step": 16909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6022710206684416,
+      "learning_rate": 5.39861731475611e-06,
+      "loss": 1.2695,
+      "step": 16910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6303610779377022,
+      "learning_rate": 5.396028964192651e-06,
+      "loss": 0.5796,
+      "step": 16911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.319526924702308,
+      "learning_rate": 5.393441159199722e-06,
+      "loss": 1.1167,
+      "step": 16912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.0312663233226935,
+      "learning_rate": 5.390853899849344e-06,
+      "loss": 1.1572,
+      "step": 16913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.706987036582027,
+      "learning_rate": 5.388267186213511e-06,
+      "loss": 1.3652,
+      "step": 16914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.395106882724713,
+      "learning_rate": 5.385681018364228e-06,
+      "loss": 1.3164,
+      "step": 16915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.932676059308892,
+      "learning_rate": 5.383095396373447e-06,
+      "loss": 0.933,
+      "step": 16916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.941107289313483,
+      "learning_rate": 5.380510320313145e-06,
+      "loss": 0.8193,
+      "step": 16917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.9088415602733795,
+      "learning_rate": 5.377925790255253e-06,
+      "loss": 1.3164,
+      "step": 16918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.913764345190214,
+      "learning_rate": 5.375341806271691e-06,
+      "loss": 0.7786,
+      "step": 16919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.396027000614195,
+      "learning_rate": 5.372758368434383e-06,
+      "loss": 1.2754,
+      "step": 16920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2524220887320623,
+      "learning_rate": 5.37017547681522e-06,
+      "loss": 0.6287,
+      "step": 16921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1504843142745127,
+      "learning_rate": 5.367593131486084e-06,
+      "loss": 0.8691,
+      "step": 16922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3191447280080855,
+      "learning_rate": 5.365011332518829e-06,
+      "loss": 1.4512,
+      "step": 16923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.230833862137426,
+      "learning_rate": 5.3624300799853255e-06,
+      "loss": 1.0898,
+      "step": 16924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0248206871609717,
+      "learning_rate": 5.359849373957387e-06,
+      "loss": 1.1465,
+      "step": 16925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9908208609959819,
+      "learning_rate": 5.357269214506855e-06,
+      "loss": 0.3824,
+      "step": 16926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.211070810658298,
+      "learning_rate": 5.354689601705518e-06,
+      "loss": 0.9746,
+      "step": 16927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.625294986680515,
+      "learning_rate": 5.352110535625177e-06,
+      "loss": 1.1055,
+      "step": 16928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2211541059292523,
+      "learning_rate": 5.349532016337602e-06,
+      "loss": 1.2578,
+      "step": 16929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6583863641256507,
+      "learning_rate": 5.346954043914551e-06,
+      "loss": 0.8545,
+      "step": 16930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.057841038487924,
+      "learning_rate": 5.34437661842776e-06,
+      "loss": 1.0645,
+      "step": 16931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.841186561656132,
+      "learning_rate": 5.3417997399489745e-06,
+      "loss": 1.5156,
+      "step": 16932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4282458772360185,
+      "learning_rate": 5.339223408549898e-06,
+      "loss": 0.9077,
+      "step": 16933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7166681320872965,
+      "learning_rate": 5.336647624302232e-06,
+      "loss": 1.2734,
+      "step": 16934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.88118398121879,
+      "learning_rate": 5.33407238727765e-06,
+      "loss": 0.917,
+      "step": 16935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5054787683479933,
+      "learning_rate": 5.331497697547827e-06,
+      "loss": 1.3516,
+      "step": 16936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5084720112769574,
+      "learning_rate": 5.328923555184423e-06,
+      "loss": 1.3809,
+      "step": 16937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.993205412960747,
+      "learning_rate": 5.326349960259067e-06,
+      "loss": 1.1387,
+      "step": 16938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.512743601471512,
+      "learning_rate": 5.323776912843384e-06,
+      "loss": 1.0547,
+      "step": 16939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.11208123854901,
+      "learning_rate": 5.321204413008973e-06,
+      "loss": 0.8184,
+      "step": 16940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.339926985210983,
+      "learning_rate": 5.318632460827438e-06,
+      "loss": 1.3301,
+      "step": 16941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.162964542978307,
+      "learning_rate": 5.31606105637035e-06,
+      "loss": 1.1514,
+      "step": 16942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6152163051192794,
+      "learning_rate": 5.313490199709267e-06,
+      "loss": 1.2695,
+      "step": 16943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.043377777697139,
+      "learning_rate": 5.310919890915725e-06,
+      "loss": 0.7285,
+      "step": 16944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4948043238903432,
+      "learning_rate": 5.308350130061282e-06,
+      "loss": 1.4082,
+      "step": 16945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.019496507925941,
+      "learning_rate": 5.305780917217437e-06,
+      "loss": 1.0986,
+      "step": 16946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6501434321628545,
+      "learning_rate": 5.303212252455695e-06,
+      "loss": 0.9492,
+      "step": 16947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.030786684256529,
+      "learning_rate": 5.300644135847527e-06,
+      "loss": 1.207,
+      "step": 16948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1037258031411192,
+      "learning_rate": 5.298076567464419e-06,
+      "loss": 1.1309,
+      "step": 16949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.56468321175199,
+      "learning_rate": 5.295509547377822e-06,
+      "loss": 0.9639,
+      "step": 16950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.692582757752978,
+      "learning_rate": 5.292943075659171e-06,
+      "loss": 1.2891,
+      "step": 16951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.048577561673006,
+      "learning_rate": 5.290377152379885e-06,
+      "loss": 0.8486,
+      "step": 16952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.809506604584103,
+      "learning_rate": 5.287811777611382e-06,
+      "loss": 1.4941,
+      "step": 16953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.837371761386686,
+      "learning_rate": 5.285246951425054e-06,
+      "loss": 0.8577,
+      "step": 16954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.486073667345704,
+      "learning_rate": 5.282682673892281e-06,
+      "loss": 1.5273,
+      "step": 16955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0761391677328667,
+      "learning_rate": 5.280118945084422e-06,
+      "loss": 1.1895,
+      "step": 16956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.470769367681774,
+      "learning_rate": 5.2775557650728165e-06,
+      "loss": 0.998,
+      "step": 16957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8988305002790598,
+      "learning_rate": 5.274993133928813e-06,
+      "loss": 0.8687,
+      "step": 16958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.217691043615071,
+      "learning_rate": 5.2724310517237225e-06,
+      "loss": 1.3323,
+      "step": 16959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.712905265702986,
+      "learning_rate": 5.2698695185288415e-06,
+      "loss": 0.8682,
+      "step": 16960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.416155131724987,
+      "learning_rate": 5.267308534415455e-06,
+      "loss": 0.9258,
+      "step": 16961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.6877463721061465,
+      "learning_rate": 5.264748099454845e-06,
+      "loss": 0.8801,
+      "step": 16962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.143353581824798,
+      "learning_rate": 5.262188213718252e-06,
+      "loss": 0.7871,
+      "step": 16963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.553761352274828,
+      "learning_rate": 5.259628877276934e-06,
+      "loss": 1.1426,
+      "step": 16964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.007056361683517,
+      "learning_rate": 5.257070090202102e-06,
+      "loss": 0.645,
+      "step": 16965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5079812441707094,
+      "learning_rate": 5.254511852564977e-06,
+      "loss": 1.4863,
+      "step": 16966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.164522968976661,
+      "learning_rate": 5.2519541644367505e-06,
+      "loss": 1.3008,
+      "step": 16967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.9289368361707515,
+      "learning_rate": 5.2493970258886e-06,
+      "loss": 1.3301,
+      "step": 16968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.968259720078525,
+      "learning_rate": 5.246840436991684e-06,
+      "loss": 0.7473,
+      "step": 16969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.211046433322558,
+      "learning_rate": 5.2442843978171614e-06,
+      "loss": 1.3477,
+      "step": 16970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.060426857969603,
+      "learning_rate": 5.241728908436164e-06,
+      "loss": 1.0017,
+      "step": 16971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.987976457597554,
+      "learning_rate": 5.2391739689198e-06,
+      "loss": 0.5396,
+      "step": 16972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2421307294540296,
+      "learning_rate": 5.236619579339186e-06,
+      "loss": 1.0088,
+      "step": 16973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1949068665399714,
+      "learning_rate": 5.234065739765398e-06,
+      "loss": 0.9163,
+      "step": 16974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.545335074059082,
+      "learning_rate": 5.231512450269518e-06,
+      "loss": 1.2383,
+      "step": 16975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.47088665789422,
+      "learning_rate": 5.228959710922601e-06,
+      "loss": 1.4941,
+      "step": 16976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9381296303800488,
+      "learning_rate": 5.226407521795684e-06,
+      "loss": 0.8896,
+      "step": 16977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.428928759677027,
+      "learning_rate": 5.223855882959791e-06,
+      "loss": 0.8721,
+      "step": 16978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.226735877252109,
+      "learning_rate": 5.221304794485946e-06,
+      "loss": 1.7305,
+      "step": 16979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.412832693769665,
+      "learning_rate": 5.218754256445135e-06,
+      "loss": 1.0068,
+      "step": 16980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1080504603345362,
+      "learning_rate": 5.216204268908331e-06,
+      "loss": 1.1934,
+      "step": 16981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9474393819928615,
+      "learning_rate": 5.213654831946513e-06,
+      "loss": 1.1582,
+      "step": 16982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.297847408441594,
+      "learning_rate": 5.2111059456306286e-06,
+      "loss": 1.4229,
+      "step": 16983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9493855192898024,
+      "learning_rate": 5.208557610031611e-06,
+      "loss": 0.6885,
+      "step": 16984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0399617333262423,
+      "learning_rate": 5.206009825220379e-06,
+      "loss": 0.8369,
+      "step": 16985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6521086102031814,
+      "learning_rate": 5.203462591267827e-06,
+      "loss": 0.7734,
+      "step": 16986
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3671271691015012,
+      "learning_rate": 5.2009159082448604e-06,
+      "loss": 0.6646,
+      "step": 16987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0333271578491843,
+      "learning_rate": 5.198369776222345e-06,
+      "loss": 1.376,
+      "step": 16988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.080793070663017,
+      "learning_rate": 5.1958241952711345e-06,
+      "loss": 0.748,
+      "step": 16989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9632091965908416,
+      "learning_rate": 5.193279165462067e-06,
+      "loss": 0.5713,
+      "step": 16990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.784390972657411,
+      "learning_rate": 5.190734686865981e-06,
+      "loss": 1.6162,
+      "step": 16991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3914785419692186,
+      "learning_rate": 5.188190759553687e-06,
+      "loss": 1.4463,
+      "step": 16992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.67799695896606,
+      "learning_rate": 5.185647383595982e-06,
+      "loss": 0.7639,
+      "step": 16993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.02193522033334,
+      "learning_rate": 5.183104559063642e-06,
+      "loss": 0.8682,
+      "step": 16994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.799199339529158,
+      "learning_rate": 5.180562286027429e-06,
+      "loss": 0.8857,
+      "step": 16995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.924540710153423,
+      "learning_rate": 5.178020564558106e-06,
+      "loss": 0.3696,
+      "step": 16996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3801336520197873,
+      "learning_rate": 5.175479394726399e-06,
+      "loss": 0.5219,
+      "step": 16997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.537320031338428,
+      "learning_rate": 5.172938776603032e-06,
+      "loss": 1.2334,
+      "step": 16998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.385995007041345,
+      "learning_rate": 5.170398710258698e-06,
+      "loss": 2.8379,
+      "step": 16999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8807421179295296,
+      "learning_rate": 5.1678591957640995e-06,
+      "loss": 0.7236,
+      "step": 17000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.682654603767323,
+      "learning_rate": 5.165320233189908e-06,
+      "loss": 1.1807,
+      "step": 17001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.89907936741802,
+      "learning_rate": 5.1627818226067825e-06,
+      "loss": 1.1406,
+      "step": 17002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.458729943647943,
+      "learning_rate": 5.160243964085356e-06,
+      "loss": 0.7646,
+      "step": 17003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2751426898651377,
+      "learning_rate": 5.157706657696271e-06,
+      "loss": 0.9766,
+      "step": 17004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0060426503874087,
+      "learning_rate": 5.155169903510129e-06,
+      "loss": 0.8291,
+      "step": 17005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8419512363444703,
+      "learning_rate": 5.1526337015975315e-06,
+      "loss": 1.0547,
+      "step": 17006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6024509087208463,
+      "learning_rate": 5.150098052029051e-06,
+      "loss": 1.125,
+      "step": 17007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.053631064929211,
+      "learning_rate": 5.147562954875268e-06,
+      "loss": 1.3301,
+      "step": 17008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.218763629180793,
+      "learning_rate": 5.14502841020672e-06,
+      "loss": 0.7432,
+      "step": 17009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.177358886862236,
+      "learning_rate": 5.142494418093955e-06,
+      "loss": 1.502,
+      "step": 17010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3768544235361984,
+      "learning_rate": 5.139960978607486e-06,
+      "loss": 1.6211,
+      "step": 17011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7124965263379615,
+      "learning_rate": 5.137428091817808e-06,
+      "loss": 1.1138,
+      "step": 17012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.946964680833166,
+      "learning_rate": 5.134895757795433e-06,
+      "loss": 0.7568,
+      "step": 17013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4416459843233724,
+      "learning_rate": 5.132363976610818e-06,
+      "loss": 1.2441,
+      "step": 17014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5612626185516643,
+      "learning_rate": 5.129832748334426e-06,
+      "loss": 1.3438,
+      "step": 17015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3689402114304965,
+      "learning_rate": 5.127302073036691e-06,
+      "loss": 1.2578,
+      "step": 17016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.009775130784794,
+      "learning_rate": 5.12477195078806e-06,
+      "loss": 0.4971,
+      "step": 17017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1346442368436565,
+      "learning_rate": 5.122242381658923e-06,
+      "loss": 1.4062,
+      "step": 17018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9614883804113261,
+      "learning_rate": 5.1197133657197e-06,
+      "loss": 0.7622,
+      "step": 17019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5521351607892493,
+      "learning_rate": 5.117184903040751e-06,
+      "loss": 1.0205,
+      "step": 17020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7011498969584458,
+      "learning_rate": 5.114656993692457e-06,
+      "loss": 0.8408,
+      "step": 17021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.543651788586768,
+      "learning_rate": 5.112129637745167e-06,
+      "loss": 1.0605,
+      "step": 17022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8847241332868188,
+      "learning_rate": 5.109602835269211e-06,
+      "loss": 0.7319,
+      "step": 17023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.319507497766723,
+      "learning_rate": 5.107076586334905e-06,
+      "loss": 0.959,
+      "step": 17024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.898074079979227,
+      "learning_rate": 5.104550891012566e-06,
+      "loss": 0.6281,
+      "step": 17025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.059648548712447,
+      "learning_rate": 5.102025749372474e-06,
+      "loss": 0.9844,
+      "step": 17026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.645812174417768,
+      "learning_rate": 5.099501161484901e-06,
+      "loss": 0.646,
+      "step": 17027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.256653010128123,
+      "learning_rate": 5.096977127420113e-06,
+      "loss": 0.9766,
+      "step": 17028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.749633113932072,
+      "learning_rate": 5.094453647248346e-06,
+      "loss": 0.9349,
+      "step": 17029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9431296438009196,
+      "learning_rate": 5.091930721039834e-06,
+      "loss": 0.9502,
+      "step": 17030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0274009256193533,
+      "learning_rate": 5.089408348864788e-06,
+      "loss": 0.7705,
+      "step": 17031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.441941182611679,
+      "learning_rate": 5.0868865307934e-06,
+      "loss": 1.165,
+      "step": 17032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9700852362109618,
+      "learning_rate": 5.084365266895847e-06,
+      "loss": 1.502,
+      "step": 17033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.183367155832493,
+      "learning_rate": 5.081844557242307e-06,
+      "loss": 0.7822,
+      "step": 17034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.605491992669474,
+      "learning_rate": 5.079324401902927e-06,
+      "loss": 0.6865,
+      "step": 17035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.441574017673201,
+      "learning_rate": 5.076804800947835e-06,
+      "loss": 1.2695,
+      "step": 17036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3360569135578326,
+      "learning_rate": 5.07428575444715e-06,
+      "loss": 0.9795,
+      "step": 17037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.327876665723715,
+      "learning_rate": 5.071767262470984e-06,
+      "loss": 0.7383,
+      "step": 17038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.820691673012819,
+      "learning_rate": 5.069249325089423e-06,
+      "loss": 0.8301,
+      "step": 17039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9832939326892784,
+      "learning_rate": 5.066731942372543e-06,
+      "loss": 1.4277,
+      "step": 17040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.017671829299559,
+      "learning_rate": 5.064215114390392e-06,
+      "loss": 1.0791,
+      "step": 17041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.889322183181145,
+      "learning_rate": 5.0616988412130225e-06,
+      "loss": 0.6597,
+      "step": 17042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.355635148429113,
+      "learning_rate": 5.059183122910461e-06,
+      "loss": 1.0234,
+      "step": 17043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8744302519782512,
+      "learning_rate": 5.056667959552713e-06,
+      "loss": 1.0715,
+      "step": 17044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.187821391882697,
+      "learning_rate": 5.05415335120977e-06,
+      "loss": 1.3633,
+      "step": 17045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.897945884358911,
+      "learning_rate": 5.051639297951618e-06,
+      "loss": 1.3926,
+      "step": 17046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.012865109357548,
+      "learning_rate": 5.049125799848234e-06,
+      "loss": 1.2949,
+      "step": 17047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9753928301816694,
+      "learning_rate": 5.046612856969554e-06,
+      "loss": 1.4688,
+      "step": 17048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4250363591624065,
+      "learning_rate": 5.044100469385516e-06,
+      "loss": 0.8467,
+      "step": 17049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.753265864324562,
+      "learning_rate": 5.041588637166031e-06,
+      "loss": 0.9043,
+      "step": 17050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4825036550287025,
+      "learning_rate": 5.039077360381015e-06,
+      "loss": 0.1842,
+      "step": 17051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6784885938642593,
+      "learning_rate": 5.036566639100351e-06,
+      "loss": 1.0762,
+      "step": 17052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0409868886284457,
+      "learning_rate": 5.034056473393911e-06,
+      "loss": 1.1602,
+      "step": 17053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.390998063243431,
+      "learning_rate": 5.031546863331543e-06,
+      "loss": 1.0742,
+      "step": 17054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.3560329343927995,
+      "learning_rate": 5.029037808983097e-06,
+      "loss": 1.2402,
+      "step": 17055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.197079078785519,
+      "learning_rate": 5.026529310418407e-06,
+      "loss": 1.1436,
+      "step": 17056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7920370188393084,
+      "learning_rate": 5.0240213677072714e-06,
+      "loss": 1.3301,
+      "step": 17057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.726349830571925,
+      "learning_rate": 5.021513980919493e-06,
+      "loss": 1.7988,
+      "step": 17058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.885404584471451,
+      "learning_rate": 5.019007150124838e-06,
+      "loss": 1.1797,
+      "step": 17059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2127959356297473,
+      "learning_rate": 5.016500875393088e-06,
+      "loss": 1.4766,
+      "step": 17060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.53256643466103,
+      "learning_rate": 5.013995156793983e-06,
+      "loss": 1.4473,
+      "step": 17061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.255821003541886,
+      "learning_rate": 5.0114899943972476e-06,
+      "loss": 0.8936,
+      "step": 17062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.213846833379911,
+      "learning_rate": 5.0089853882726165e-06,
+      "loss": 1.7012,
+      "step": 17063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.796337184516247,
+      "learning_rate": 5.006481338489777e-06,
+      "loss": 1.0332,
+      "step": 17064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.281875106656885,
+      "learning_rate": 5.0039778451184295e-06,
+      "loss": 1.1227,
+      "step": 17065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4934014021749817,
+      "learning_rate": 5.001474908228238e-06,
+      "loss": 1.2383,
+      "step": 17066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.595912408826273,
+      "learning_rate": 4.998972527888854e-06,
+      "loss": 1.123,
+      "step": 17067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0305694130653444,
+      "learning_rate": 4.996470704169926e-06,
+      "loss": 1.0962,
+      "step": 17068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1565695263281466,
+      "learning_rate": 4.993969437141074e-06,
+      "loss": 1.0347,
+      "step": 17069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3528034573109013,
+      "learning_rate": 4.991468726871912e-06,
+      "loss": 1.4414,
+      "step": 17070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0965792887185692,
+      "learning_rate": 4.9889685734320175e-06,
+      "loss": 0.6496,
+      "step": 17071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.236943189504116,
+      "learning_rate": 4.986468976890993e-06,
+      "loss": 0.6111,
+      "step": 17072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.281902829686328,
+      "learning_rate": 4.983969937318386e-06,
+      "loss": 0.9141,
+      "step": 17073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.562953769641091,
+      "learning_rate": 4.981471454783742e-06,
+      "loss": 1.75,
+      "step": 17074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.884718981463319,
+      "learning_rate": 4.978973529356606e-06,
+      "loss": 0.6318,
+      "step": 17075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4560231383100417,
+      "learning_rate": 4.976476161106478e-06,
+      "loss": 1.0049,
+      "step": 17076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6326263995858095,
+      "learning_rate": 4.973979350102872e-06,
+      "loss": 1.0684,
+      "step": 17077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6965010966180785,
+      "learning_rate": 4.971483096415269e-06,
+      "loss": 1.0625,
+      "step": 17078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.436140219112047,
+      "learning_rate": 4.96898740011314e-06,
+      "loss": 1.1426,
+      "step": 17079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.950325905852543,
+      "learning_rate": 4.966492261265929e-06,
+      "loss": 0.4604,
+      "step": 17080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.057309949431377,
+      "learning_rate": 4.963997679943089e-06,
+      "loss": 0.835,
+      "step": 17081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.690638107285601,
+      "learning_rate": 4.961503656214036e-06,
+      "loss": 1.1113,
+      "step": 17082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4264597378553923,
+      "learning_rate": 4.9590101901481725e-06,
+      "loss": 1.1953,
+      "step": 17083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.717536151148394,
+      "learning_rate": 4.9565172818148984e-06,
+      "loss": 1.2988,
+      "step": 17084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9083752900998716,
+      "learning_rate": 4.954024931283594e-06,
+      "loss": 0.9404,
+      "step": 17085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7317098812381102,
+      "learning_rate": 4.951533138623615e-06,
+      "loss": 1.2988,
+      "step": 17086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4735278481172145,
+      "learning_rate": 4.9490419039043055e-06,
+      "loss": 0.895,
+      "step": 17087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.184822453890511,
+      "learning_rate": 4.9465512271949925e-06,
+      "loss": 0.7593,
+      "step": 17088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.136485850047602,
+      "learning_rate": 4.944061108565001e-06,
+      "loss": 1.1758,
+      "step": 17089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.77738289463448,
+      "learning_rate": 4.941571548083623e-06,
+      "loss": 1.6143,
+      "step": 17090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.227949493472661,
+      "learning_rate": 4.939082545820145e-06,
+      "loss": 1.0977,
+      "step": 17091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3791628043103863,
+      "learning_rate": 4.936594101843828e-06,
+      "loss": 1.1025,
+      "step": 17092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8799569892320718,
+      "learning_rate": 4.934106216223925e-06,
+      "loss": 0.7854,
+      "step": 17093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.804906044948741,
+      "learning_rate": 4.931618889029685e-06,
+      "loss": 0.5474,
+      "step": 17094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.841164083079586,
+      "learning_rate": 4.929132120330324e-06,
+      "loss": 0.7183,
+      "step": 17095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.100415111089602,
+      "learning_rate": 4.926645910195043e-06,
+      "loss": 1.2002,
+      "step": 17096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.354288745066693,
+      "learning_rate": 4.924160258693028e-06,
+      "loss": 0.9355,
+      "step": 17097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.403115530660395,
+      "learning_rate": 4.921675165893466e-06,
+      "loss": 0.9658,
+      "step": 17098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8470567195107526,
+      "learning_rate": 4.919190631865511e-06,
+      "loss": 0.6294,
+      "step": 17099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.255007094344827,
+      "learning_rate": 4.916706656678302e-06,
+      "loss": 1.5449,
+      "step": 17100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.747648670585087,
+      "learning_rate": 4.914223240400973e-06,
+      "loss": 0.6997,
+      "step": 17101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0382179322318783,
+      "learning_rate": 4.911740383102631e-06,
+      "loss": 1.0625,
+      "step": 17102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6928073906648047,
+      "learning_rate": 4.909258084852381e-06,
+      "loss": 1.2383,
+      "step": 17103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.526908952374229,
+      "learning_rate": 4.9067763457193e-06,
+      "loss": 1.541,
+      "step": 17104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.782641202183478,
+      "learning_rate": 4.9042951657724465e-06,
+      "loss": 0.7385,
+      "step": 17105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.501617480596519,
+      "learning_rate": 4.901814545080885e-06,
+      "loss": 0.9785,
+      "step": 17106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.684504994770819,
+      "learning_rate": 4.899334483713641e-06,
+      "loss": 0.645,
+      "step": 17107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0746061893554315,
+      "learning_rate": 4.896854981739737e-06,
+      "loss": 0.7715,
+      "step": 17108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.756046323971037,
+      "learning_rate": 4.894376039228169e-06,
+      "loss": 0.6917,
+      "step": 17109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.032583884447196,
+      "learning_rate": 4.891897656247935e-06,
+      "loss": 1.1953,
+      "step": 17110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.056034876177138,
+      "learning_rate": 4.889419832867997e-06,
+      "loss": 1.543,
+      "step": 17111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8866569708587173,
+      "learning_rate": 4.8869425691573265e-06,
+      "loss": 1.2305,
+      "step": 17112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.316480252658289,
+      "learning_rate": 4.884465865184854e-06,
+      "loss": 1.123,
+      "step": 17113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6738148390089473,
+      "learning_rate": 4.881989721019503e-06,
+      "loss": 0.9102,
+      "step": 17114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.221092421527762,
+      "learning_rate": 4.8795141367301924e-06,
+      "loss": 0.571,
+      "step": 17115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.71130139989395,
+      "learning_rate": 4.877039112385815e-06,
+      "loss": 1.1104,
+      "step": 17116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6959598372040223,
+      "learning_rate": 4.874564648055244e-06,
+      "loss": 0.9404,
+      "step": 17117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.039664306668475,
+      "learning_rate": 4.872090743807342e-06,
+      "loss": 0.7627,
+      "step": 17118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.1557302502014,
+      "learning_rate": 4.869617399710966e-06,
+      "loss": 0.8858,
+      "step": 17119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.170254749206398,
+      "learning_rate": 4.867144615834934e-06,
+      "loss": 1.4863,
+      "step": 17120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.1402090187566785,
+      "learning_rate": 4.8646723922480815e-06,
+      "loss": 1.2246,
+      "step": 17121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.2765100640879,
+      "learning_rate": 4.86220072901919e-06,
+      "loss": 1.0381,
+      "step": 17122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.756871560770887,
+      "learning_rate": 4.859729626217063e-06,
+      "loss": 0.6951,
+      "step": 17123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.022425124737345,
+      "learning_rate": 4.85725908391046e-06,
+      "loss": 1.0361,
+      "step": 17124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9006794692557496,
+      "learning_rate": 4.854789102168136e-06,
+      "loss": 1.1035,
+      "step": 17125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.607941255661753,
+      "learning_rate": 4.852319681058823e-06,
+      "loss": 1.1973,
+      "step": 17126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.239403568120732,
+      "learning_rate": 4.8498508206512614e-06,
+      "loss": 1.5957,
+      "step": 17127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8805241271147048,
+      "learning_rate": 4.847382521014143e-06,
+      "loss": 0.918,
+      "step": 17128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.532589875685377,
+      "learning_rate": 4.8449147822161616e-06,
+      "loss": 0.9883,
+      "step": 17129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.878421995239396,
+      "learning_rate": 4.842447604326003e-06,
+      "loss": 1.0996,
+      "step": 17130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.719036824688965,
+      "learning_rate": 4.839980987412315e-06,
+      "loss": 1.165,
+      "step": 17131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.972906137856677,
+      "learning_rate": 4.837514931543754e-06,
+      "loss": 0.9943,
+      "step": 17132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8745124182932076,
+      "learning_rate": 4.8350494367889464e-06,
+      "loss": 0.3794,
+      "step": 17133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5478898349283856,
+      "learning_rate": 4.832584503216503e-06,
+      "loss": 0.7651,
+      "step": 17134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.061031134693888,
+      "learning_rate": 4.830120130895016e-06,
+      "loss": 0.8262,
+      "step": 17135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8069397450140934,
+      "learning_rate": 4.827656319893081e-06,
+      "loss": 1.0898,
+      "step": 17136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9440823118375126,
+      "learning_rate": 4.8251930702792605e-06,
+      "loss": 1.2686,
+      "step": 17137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7628437518653208,
+      "learning_rate": 4.8227303821221005e-06,
+      "loss": 0.8418,
+      "step": 17138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.32244015126883,
+      "learning_rate": 4.820268255490129e-06,
+      "loss": 1.1699,
+      "step": 17139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.729060153492195,
+      "learning_rate": 4.81780669045189e-06,
+      "loss": 1.0283,
+      "step": 17140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.510293464222814,
+      "learning_rate": 4.815345687075873e-06,
+      "loss": 1.5137,
+      "step": 17141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.360218472387952,
+      "learning_rate": 4.812885245430568e-06,
+      "loss": 1.3369,
+      "step": 17142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3576599354866525,
+      "learning_rate": 4.810425365584445e-06,
+      "loss": 1.3262,
+      "step": 17143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5764781894156736,
+      "learning_rate": 4.807966047605969e-06,
+      "loss": 1.0146,
+      "step": 17144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.466338319661609,
+      "learning_rate": 4.805507291563577e-06,
+      "loss": 0.8604,
+      "step": 17145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.505612082406457,
+      "learning_rate": 4.803049097525697e-06,
+      "loss": 1.1936,
+      "step": 17146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9438485148202074,
+      "learning_rate": 4.800591465560733e-06,
+      "loss": 0.4863,
+      "step": 17147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2380978174590886,
+      "learning_rate": 4.798134395737083e-06,
+      "loss": 1.2178,
+      "step": 17148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3192795017804886,
+      "learning_rate": 4.795677888123137e-06,
+      "loss": 1.0312,
+      "step": 17149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.631142326472689,
+      "learning_rate": 4.793221942787249e-06,
+      "loss": 1.3926,
+      "step": 17150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2826847391082974,
+      "learning_rate": 4.790766559797769e-06,
+      "loss": 1.125,
+      "step": 17151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.034103304937872,
+      "learning_rate": 4.788311739223019e-06,
+      "loss": 2.377,
+      "step": 17152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.246063603737911,
+      "learning_rate": 4.785857481131336e-06,
+      "loss": 0.8818,
+      "step": 17153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1356984270493973,
+      "learning_rate": 4.783403785591009e-06,
+      "loss": 0.7949,
+      "step": 17154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0417613442195606,
+      "learning_rate": 4.780950652670324e-06,
+      "loss": 0.4766,
+      "step": 17155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.289008468537743,
+      "learning_rate": 4.778498082437544e-06,
+      "loss": 0.9766,
+      "step": 17156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3958156308612018,
+      "learning_rate": 4.77604607496093e-06,
+      "loss": 1.2227,
+      "step": 17157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.778008829481107,
+      "learning_rate": 4.773594630308728e-06,
+      "loss": 0.6196,
+      "step": 17158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8499279291703203,
+      "learning_rate": 4.771143748549153e-06,
+      "loss": 0.8147,
+      "step": 17159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.591978084106536,
+      "learning_rate": 4.768693429750406e-06,
+      "loss": 0.9717,
+      "step": 17160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5337473476642565,
+      "learning_rate": 4.766243673980692e-06,
+      "loss": 1.6895,
+      "step": 17161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.585862069977714,
+      "learning_rate": 4.763794481308176e-06,
+      "loss": 0.8945,
+      "step": 17162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.961303068941913,
+      "learning_rate": 4.761345851801025e-06,
+      "loss": 0.8518,
+      "step": 17163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.794722341349938,
+      "learning_rate": 4.758897785527372e-06,
+      "loss": 1.2598,
+      "step": 17164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.063686090620233,
+      "learning_rate": 4.756450282555358e-06,
+      "loss": 1.2051,
+      "step": 17165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7503267440886487,
+      "learning_rate": 4.754003342953089e-06,
+      "loss": 0.9141,
+      "step": 17166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0070124367144984,
+      "learning_rate": 4.751556966788668e-06,
+      "loss": 0.9971,
+      "step": 17167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9271421234272497,
+      "learning_rate": 4.749111154130176e-06,
+      "loss": 0.5776,
+      "step": 17168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3223727035539854,
+      "learning_rate": 4.746665905045669e-06,
+      "loss": 1.3086,
+      "step": 17169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.001654521177503,
+      "learning_rate": 4.74422121960321e-06,
+      "loss": 0.9038,
+      "step": 17170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.808376425875645,
+      "learning_rate": 4.741777097870828e-06,
+      "loss": 1.3535,
+      "step": 17171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.962604621427447,
+      "learning_rate": 4.739333539916543e-06,
+      "loss": 1.209,
+      "step": 17172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.355778865105827,
+      "learning_rate": 4.7368905458083534e-06,
+      "loss": 1.4531,
+      "step": 17173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.298653085586197,
+      "learning_rate": 4.734448115614254e-06,
+      "loss": 0.8232,
+      "step": 17174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.316683927878789,
+      "learning_rate": 4.732006249402215e-06,
+      "loss": 1.1641,
+      "step": 17175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.244887087527537,
+      "learning_rate": 4.729564947240187e-06,
+      "loss": 0.9199,
+      "step": 17176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.428483278213432,
+      "learning_rate": 4.72712420919611e-06,
+      "loss": 0.8252,
+      "step": 17177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8178528603502753,
+      "learning_rate": 4.72468403533792e-06,
+      "loss": 1.4824,
+      "step": 17178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.41361972198475,
+      "learning_rate": 4.722244425733521e-06,
+      "loss": 0.9961,
+      "step": 17179
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.720363620407001,
+      "learning_rate": 4.719805380450803e-06,
+      "loss": 1.125,
+      "step": 17180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0175251878952922,
+      "learning_rate": 4.71736689955764e-06,
+      "loss": 1.0627,
+      "step": 17181
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.76278051316169,
+      "learning_rate": 4.714928983121908e-06,
+      "loss": 1.1406,
+      "step": 17182
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2922726147950967,
+      "learning_rate": 4.712491631211441e-06,
+      "loss": 1.1035,
+      "step": 17183
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.58166953826935,
+      "learning_rate": 4.7100548438940736e-06,
+      "loss": 1.0039,
+      "step": 17184
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7503521433693945,
+      "learning_rate": 4.707618621237614e-06,
+      "loss": 0.9509,
+      "step": 17185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.104452559514302,
+      "learning_rate": 4.7051829633098646e-06,
+      "loss": 0.9854,
+      "step": 17186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.146660227368004,
+      "learning_rate": 4.70274787017862e-06,
+      "loss": 0.9365,
+      "step": 17187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3455724295382727,
+      "learning_rate": 4.700313341911638e-06,
+      "loss": 0.7261,
+      "step": 17188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.199424280590789,
+      "learning_rate": 4.697879378576675e-06,
+      "loss": 0.9316,
+      "step": 17189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.976814288462967,
+      "learning_rate": 4.6954459802414555e-06,
+      "loss": 1.2021,
+      "step": 17190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.172086996952427,
+      "learning_rate": 4.6930131469737125e-06,
+      "loss": 1.6387,
+      "step": 17191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.751138185606813,
+      "learning_rate": 4.69058087884115e-06,
+      "loss": 2.1074,
+      "step": 17192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3804979512311024,
+      "learning_rate": 4.688149175911452e-06,
+      "loss": 0.4185,
+      "step": 17193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.526751791406931,
+      "learning_rate": 4.685718038252287e-06,
+      "loss": 0.8057,
+      "step": 17194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7566814750940947,
+      "learning_rate": 4.683287465931319e-06,
+      "loss": 1.3379,
+      "step": 17195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1194991551186293,
+      "learning_rate": 4.680857459016197e-06,
+      "loss": 1.0039,
+      "step": 17196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.212960972349161,
+      "learning_rate": 4.678428017574538e-06,
+      "loss": 0.8291,
+      "step": 17197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.233992323750394,
+      "learning_rate": 4.675999141673951e-06,
+      "loss": 0.9727,
+      "step": 17198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.534724867150526,
+      "learning_rate": 4.6735708313820375e-06,
+      "loss": 0.5719,
+      "step": 17199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.307947368464677,
+      "learning_rate": 4.671143086766375e-06,
+      "loss": 1.1211,
+      "step": 17200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.295305403457588,
+      "learning_rate": 4.6687159078945245e-06,
+      "loss": 0.8975,
+      "step": 17201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2075070064286786,
+      "learning_rate": 4.666289294834025e-06,
+      "loss": 0.7397,
+      "step": 17202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6950648670413173,
+      "learning_rate": 4.663863247652425e-06,
+      "loss": 0.7744,
+      "step": 17203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.040624378984472,
+      "learning_rate": 4.661437766417226e-06,
+      "loss": 1.2324,
+      "step": 17204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.721032043407889,
+      "learning_rate": 4.659012851195937e-06,
+      "loss": 1.3774,
+      "step": 17205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.738798578036654,
+      "learning_rate": 4.656588502056042e-06,
+      "loss": 0.8418,
+      "step": 17206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.697147110525988,
+      "learning_rate": 4.654164719065002e-06,
+      "loss": 0.8896,
+      "step": 17207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.583365920754045,
+      "learning_rate": 4.65174150229028e-06,
+      "loss": 0.7656,
+      "step": 17208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.978955231118417,
+      "learning_rate": 4.649318851799309e-06,
+      "loss": 0.8242,
+      "step": 17209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.95661019259883,
+      "learning_rate": 4.6468967676595085e-06,
+      "loss": 0.6272,
+      "step": 17210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.977577817269928,
+      "learning_rate": 4.644475249938279e-06,
+      "loss": 1.834,
+      "step": 17211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.578563908255703,
+      "learning_rate": 4.642054298703022e-06,
+      "loss": 1.3389,
+      "step": 17212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.595642622109646,
+      "learning_rate": 4.639633914021102e-06,
+      "loss": 1.2871,
+      "step": 17213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6029018127822,
+      "learning_rate": 4.6372140959598885e-06,
+      "loss": 1.3809,
+      "step": 17214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.679608021158337,
+      "learning_rate": 4.634794844586712e-06,
+      "loss": 0.6826,
+      "step": 17215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8610126412574726,
+      "learning_rate": 4.632376159968901e-06,
+      "loss": 0.9521,
+      "step": 17216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.803817996774988,
+      "learning_rate": 4.62995804217378e-06,
+      "loss": 1.6055,
+      "step": 17217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.535349287316136,
+      "learning_rate": 4.627540491268628e-06,
+      "loss": 1.0547,
+      "step": 17218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.284202233081338,
+      "learning_rate": 4.625123507320728e-06,
+      "loss": 1.1279,
+      "step": 17219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2252958208232316,
+      "learning_rate": 4.622707090397349e-06,
+      "loss": 1.0415,
+      "step": 17220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0874821527250966,
+      "learning_rate": 4.620291240565738e-06,
+      "loss": 1.1621,
+      "step": 17221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2517424094175404,
+      "learning_rate": 4.617875957893119e-06,
+      "loss": 0.9961,
+      "step": 17222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2812728154433723,
+      "learning_rate": 4.6154612424467225e-06,
+      "loss": 0.7109,
+      "step": 17223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.089784653443056,
+      "learning_rate": 4.613047094293732e-06,
+      "loss": 1.0547,
+      "step": 17224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.656899549381965,
+      "learning_rate": 4.61063351350135e-06,
+      "loss": 1.0342,
+      "step": 17225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.122045033177489,
+      "learning_rate": 4.608220500136737e-06,
+      "loss": 0.791,
+      "step": 17226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9852033673324174,
+      "learning_rate": 4.605808054267044e-06,
+      "loss": 0.5654,
+      "step": 17227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.779034430487058,
+      "learning_rate": 4.603396175959404e-06,
+      "loss": 1.0273,
+      "step": 17228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8745203040349436,
+      "learning_rate": 4.6009848652809525e-06,
+      "loss": 0.751,
+      "step": 17229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.5910325935408505,
+      "learning_rate": 4.598574122298788e-06,
+      "loss": 1.2578,
+      "step": 17230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.912172278424274,
+      "learning_rate": 4.596163947079993e-06,
+      "loss": 1.6084,
+      "step": 17231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7580590502618767,
+      "learning_rate": 4.593754339691653e-06,
+      "loss": 0.9697,
+      "step": 17232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.757122266806881,
+      "learning_rate": 4.59134530020082e-06,
+      "loss": 1.3242,
+      "step": 17233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5270418579354423,
+      "learning_rate": 4.588936828674542e-06,
+      "loss": 1.0537,
+      "step": 17234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.341055809625854,
+      "learning_rate": 4.586528925179845e-06,
+      "loss": 1.7285,
+      "step": 17235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.650785225838101,
+      "learning_rate": 4.584121589783738e-06,
+      "loss": 0.8037,
+      "step": 17236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9597406930506414,
+      "learning_rate": 4.581714822553207e-06,
+      "loss": 0.9961,
+      "step": 17237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1147495323475565,
+      "learning_rate": 4.579308623555248e-06,
+      "loss": 1.2715,
+      "step": 17238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.004951749932465,
+      "learning_rate": 4.576902992856819e-06,
+      "loss": 0.6045,
+      "step": 17239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.71730808545117,
+      "learning_rate": 4.574497930524857e-06,
+      "loss": 0.916,
+      "step": 17240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7205510805561586,
+      "learning_rate": 4.572093436626301e-06,
+      "loss": 1.2559,
+      "step": 17241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3023959391946,
+      "learning_rate": 4.569689511228076e-06,
+      "loss": 1.502,
+      "step": 17242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0653277577168363,
+      "learning_rate": 4.5672861543970745e-06,
+      "loss": 1.4102,
+      "step": 17243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0537885430956715,
+      "learning_rate": 4.564883366200182e-06,
+      "loss": 0.6824,
+      "step": 17244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.190376031410993,
+      "learning_rate": 4.562481146704259e-06,
+      "loss": 0.748,
+      "step": 17245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.650884077763076,
+      "learning_rate": 4.560079495976175e-06,
+      "loss": 0.9819,
+      "step": 17246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.575755654723595,
+      "learning_rate": 4.557678414082753e-06,
+      "loss": 0.9492,
+      "step": 17247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9099619635943852,
+      "learning_rate": 4.555277901090821e-06,
+      "loss": 0.7051,
+      "step": 17248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.214496870500672,
+      "learning_rate": 4.552877957067175e-06,
+      "loss": 1.4238,
+      "step": 17249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.367886380305056,
+      "learning_rate": 4.550478582078613e-06,
+      "loss": 0.8682,
+      "step": 17250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0755851064884254,
+      "learning_rate": 4.548079776191911e-06,
+      "loss": 1.0859,
+      "step": 17251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9070633601214717,
+      "learning_rate": 4.545681539473823e-06,
+      "loss": 1.0488,
+      "step": 17252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.581979840730888,
+      "learning_rate": 4.543283871991091e-06,
+      "loss": 1.4463,
+      "step": 17253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8189595112891994,
+      "learning_rate": 4.540886773810438e-06,
+      "loss": 1.334,
+      "step": 17254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1752701164098456,
+      "learning_rate": 4.538490244998578e-06,
+      "loss": 1.3164,
+      "step": 17255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1429657824429036,
+      "learning_rate": 4.536094285622208e-06,
+      "loss": 0.7939,
+      "step": 17256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.302168037867629,
+      "learning_rate": 4.5336988957479955e-06,
+      "loss": 1.1738,
+      "step": 17257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9253272707285405,
+      "learning_rate": 4.5313040754426185e-06,
+      "loss": 0.8672,
+      "step": 17258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.1904835299214325,
+      "learning_rate": 4.528909824772709e-06,
+      "loss": 1.8359,
+      "step": 17259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6347560330973474,
+      "learning_rate": 4.526516143804912e-06,
+      "loss": 0.7412,
+      "step": 17260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.039736698399479,
+      "learning_rate": 4.524123032605837e-06,
+      "loss": 0.8955,
+      "step": 17261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.470444785623078,
+      "learning_rate": 4.5217304912420736e-06,
+      "loss": 1.2344,
+      "step": 17262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.741828493393033,
+      "learning_rate": 4.519338519780222e-06,
+      "loss": 1.3716,
+      "step": 17263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7337430387180692,
+      "learning_rate": 4.516947118286841e-06,
+      "loss": 0.66,
+      "step": 17264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.9123454528260115,
+      "learning_rate": 4.514556286828484e-06,
+      "loss": 1.2266,
+      "step": 17265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7436608896764463,
+      "learning_rate": 4.512166025471678e-06,
+      "loss": 0.5377,
+      "step": 17266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7088901373214354,
+      "learning_rate": 4.5097763342829585e-06,
+      "loss": 1.0093,
+      "step": 17267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.306557956201863,
+      "learning_rate": 4.507387213328815e-06,
+      "loss": 1.1646,
+      "step": 17268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0934240742270123,
+      "learning_rate": 4.50499866267575e-06,
+      "loss": 1.0933,
+      "step": 17269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.24385681467367,
+      "learning_rate": 4.502610682390229e-06,
+      "loss": 1.1504,
+      "step": 17270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.091866736011441,
+      "learning_rate": 4.5002232725387005e-06,
+      "loss": 1.1001,
+      "step": 17271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.090987931525468,
+      "learning_rate": 4.4978364331876185e-06,
+      "loss": 1.0168,
+      "step": 17272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.945795075322358,
+      "learning_rate": 4.495450164403406e-06,
+      "loss": 0.9688,
+      "step": 17273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.324063485243257,
+      "learning_rate": 4.493064466252464e-06,
+      "loss": 1.7109,
+      "step": 17274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1842534604523602,
+      "learning_rate": 4.490679338801185e-06,
+      "loss": 0.9561,
+      "step": 17275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.584413384702568,
+      "learning_rate": 4.488294782115957e-06,
+      "loss": 0.918,
+      "step": 17276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.402187518924535,
+      "learning_rate": 4.4859107962631345e-06,
+      "loss": 1.0928,
+      "step": 17277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8802242083166614,
+      "learning_rate": 4.483527381309055e-06,
+      "loss": 1.2646,
+      "step": 17278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.424540333122056,
+      "learning_rate": 4.481144537320059e-06,
+      "loss": 0.7432,
+      "step": 17279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.216728579103519,
+      "learning_rate": 4.478762264362463e-06,
+      "loss": 0.9248,
+      "step": 17280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.575001171967725,
+      "learning_rate": 4.476380562502558e-06,
+      "loss": 1.5625,
+      "step": 17281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3645538686745478,
+      "learning_rate": 4.473999431806628e-06,
+      "loss": 1.418,
+      "step": 17282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.217487739766769,
+      "learning_rate": 4.471618872340933e-06,
+      "loss": 0.71,
+      "step": 17283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.138977215258588,
+      "learning_rate": 4.469238884171731e-06,
+      "loss": 1.1504,
+      "step": 17284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.24544434820646,
+      "learning_rate": 4.466859467365253e-06,
+      "loss": 1.0322,
+      "step": 17285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5264293883845466,
+      "learning_rate": 4.4644806219877184e-06,
+      "loss": 1.2051,
+      "step": 17286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5199136133373425,
+      "learning_rate": 4.4621023481053206e-06,
+      "loss": 0.9707,
+      "step": 17287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0919003592816625,
+      "learning_rate": 4.459724645784252e-06,
+      "loss": 0.4893,
+      "step": 17288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0858989812030533,
+      "learning_rate": 4.457347515090693e-06,
+      "loss": 0.8467,
+      "step": 17289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.004163580539836,
+      "learning_rate": 4.45497095609079e-06,
+      "loss": 1.0625,
+      "step": 17290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.821288652936783,
+      "learning_rate": 4.452594968850682e-06,
+      "loss": 0.6025,
+      "step": 17291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0767943318648556,
+      "learning_rate": 4.450219553436482e-06,
+      "loss": 0.5366,
+      "step": 17292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2459654132832934,
+      "learning_rate": 4.447844709914315e-06,
+      "loss": 0.8696,
+      "step": 17293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8202940710421958,
+      "learning_rate": 4.4454704383502606e-06,
+      "loss": 0.98,
+      "step": 17294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1891046360706508,
+      "learning_rate": 4.4430967388103984e-06,
+      "loss": 0.9502,
+      "step": 17295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8948280337973114,
+      "learning_rate": 4.440723611360772e-06,
+      "loss": 1.1125,
+      "step": 17296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8449550213467925,
+      "learning_rate": 4.438351056067449e-06,
+      "loss": 1.1367,
+      "step": 17297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.717197720473878,
+      "learning_rate": 4.4359790729964486e-06,
+      "loss": 0.6101,
+      "step": 17298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6672901497355377,
+      "learning_rate": 4.433607662213776e-06,
+      "loss": 0.7065,
+      "step": 17299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9386600590103091,
+      "learning_rate": 4.431236823785423e-06,
+      "loss": 1.2109,
+      "step": 17300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.22850643813998,
+      "learning_rate": 4.428866557777384e-06,
+      "loss": 1.0234,
+      "step": 17301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.981711432297016,
+      "learning_rate": 4.426496864255614e-06,
+      "loss": 1.4238,
+      "step": 17302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9973135681971335,
+      "learning_rate": 4.424127743286061e-06,
+      "loss": 1.1152,
+      "step": 17303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.947855451148217,
+      "learning_rate": 4.421759194934649e-06,
+      "loss": 0.7404,
+      "step": 17304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.9702242483581465,
+      "learning_rate": 4.4193912192673024e-06,
+      "loss": 1.3477,
+      "step": 17305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.396276346213949,
+      "learning_rate": 4.417023816349925e-06,
+      "loss": 0.7144,
+      "step": 17306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.444429535940712,
+      "learning_rate": 4.414656986248395e-06,
+      "loss": 2.2539,
+      "step": 17307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.085381489170475,
+      "learning_rate": 4.412290729028581e-06,
+      "loss": 0.7568,
+      "step": 17308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.936614856876358,
+      "learning_rate": 4.40992504475633e-06,
+      "loss": 1.1562,
+      "step": 17309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.512188405614694,
+      "learning_rate": 4.407559933497488e-06,
+      "loss": 1.4355,
+      "step": 17310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4315102170729164,
+      "learning_rate": 4.405195395317868e-06,
+      "loss": 1.0127,
+      "step": 17311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.741137619315956,
+      "learning_rate": 4.40283143028328e-06,
+      "loss": 1.1309,
+      "step": 17312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6731719931043882,
+      "learning_rate": 4.400468038459498e-06,
+      "loss": 1.0264,
+      "step": 17313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2985667728768133,
+      "learning_rate": 4.398105219912313e-06,
+      "loss": 1.125,
+      "step": 17314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.444930074117004,
+      "learning_rate": 4.395742974707468e-06,
+      "loss": 0.9326,
+      "step": 17315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.602190590498105,
+      "learning_rate": 4.39338130291071e-06,
+      "loss": 1.6152,
+      "step": 17316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4172309896234667,
+      "learning_rate": 4.391020204587756e-06,
+      "loss": 1.0547,
+      "step": 17317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.559731461758595,
+      "learning_rate": 4.38865967980433e-06,
+      "loss": 1.1753,
+      "step": 17318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.481955543734706,
+      "learning_rate": 4.386299728626111e-06,
+      "loss": 0.8838,
+      "step": 17319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.432136234976259,
+      "learning_rate": 4.383940351118779e-06,
+      "loss": 1.3296,
+      "step": 17320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.224323225350654,
+      "learning_rate": 4.381581547347988e-06,
+      "loss": 0.7168,
+      "step": 17321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1435162030286787,
+      "learning_rate": 4.379223317379396e-06,
+      "loss": 1.1992,
+      "step": 17322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7461085095287032,
+      "learning_rate": 4.376865661278623e-06,
+      "loss": 0.6909,
+      "step": 17323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.519727406517522,
+      "learning_rate": 4.374508579111277e-06,
+      "loss": 1.6035,
+      "step": 17324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7201104980351434,
+      "learning_rate": 4.372152070942967e-06,
+      "loss": 0.6694,
+      "step": 17325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9297130752907279,
+      "learning_rate": 4.369796136839263e-06,
+      "loss": 0.7988,
+      "step": 17326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.642242387757573,
+      "learning_rate": 4.367440776865739e-06,
+      "loss": 0.8457,
+      "step": 17327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.406472827452826,
+      "learning_rate": 4.365085991087936e-06,
+      "loss": 0.9298,
+      "step": 17328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6722386044994948,
+      "learning_rate": 4.3627317795713914e-06,
+      "loss": 1.3374,
+      "step": 17329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8901531797227868,
+      "learning_rate": 4.360378142381613e-06,
+      "loss": 1.3809,
+      "step": 17330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.753143876318534,
+      "learning_rate": 4.358025079584116e-06,
+      "loss": 1.4663,
+      "step": 17331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.469105658685683,
+      "learning_rate": 4.355672591244378e-06,
+      "loss": 0.9912,
+      "step": 17332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.878821762607869,
+      "learning_rate": 4.353320677427861e-06,
+      "loss": 1.4375,
+      "step": 17333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0132756544461268,
+      "learning_rate": 4.3509693382000225e-06,
+      "loss": 0.6992,
+      "step": 17334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.703647238689498,
+      "learning_rate": 4.348618573626309e-06,
+      "loss": 0.7988,
+      "step": 17335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.487426628035988,
+      "learning_rate": 4.3462683837721345e-06,
+      "loss": 1.2861,
+      "step": 17336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.444777606104004,
+      "learning_rate": 4.343918768702901e-06,
+      "loss": 0.6758,
+      "step": 17337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3201555282288235,
+      "learning_rate": 4.341569728483993e-06,
+      "loss": 0.834,
+      "step": 17338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.170437942963575,
+      "learning_rate": 4.339221263180795e-06,
+      "loss": 0.9268,
+      "step": 17339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0446896271775037,
+      "learning_rate": 4.3368733728586605e-06,
+      "loss": 1.0664,
+      "step": 17340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.272471534038734,
+      "learning_rate": 4.334526057582924e-06,
+      "loss": 0.5859,
+      "step": 17341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.406950094944805,
+      "learning_rate": 4.33217931741891e-06,
+      "loss": 1.3496,
+      "step": 17342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.180750761858278,
+      "learning_rate": 4.329833152431934e-06,
+      "loss": 0.9727,
+      "step": 17343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.866829873795334,
+      "learning_rate": 4.327487562687291e-06,
+      "loss": 1.0283,
+      "step": 17344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4626723649403197,
+      "learning_rate": 4.325142548250255e-06,
+      "loss": 0.76,
+      "step": 17345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.7247041911058245,
+      "learning_rate": 4.322798109186082e-06,
+      "loss": 1.1348,
+      "step": 17346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8016067629039187,
+      "learning_rate": 4.320454245560015e-06,
+      "loss": 1.126,
+      "step": 17347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.969625885902411,
+      "learning_rate": 4.318110957437296e-06,
+      "loss": 0.709,
+      "step": 17348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.849403085058208,
+      "learning_rate": 4.315768244883128e-06,
+      "loss": 0.6958,
+      "step": 17349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.01508069438082,
+      "learning_rate": 4.313426107962709e-06,
+      "loss": 1.0146,
+      "step": 17350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.502570717216619,
+      "learning_rate": 4.311084546741212e-06,
+      "loss": 1.5464,
+      "step": 17351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.146189120034075,
+      "learning_rate": 4.308743561283813e-06,
+      "loss": 1.0996,
+      "step": 17352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.273250819334833,
+      "learning_rate": 4.306403151655661e-06,
+      "loss": 1.5547,
+      "step": 17353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.775339597250383,
+      "learning_rate": 4.304063317921886e-06,
+      "loss": 1.6504,
+      "step": 17354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.891362448385642,
+      "learning_rate": 4.301724060147597e-06,
+      "loss": 0.9785,
+      "step": 17355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7757253746981903,
+      "learning_rate": 4.299385378397907e-06,
+      "loss": 1.3418,
+      "step": 17356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.013318183391404,
+      "learning_rate": 4.2970472727378945e-06,
+      "loss": 0.8721,
+      "step": 17357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3744545955070535,
+      "learning_rate": 4.2947097432326275e-06,
+      "loss": 0.9902,
+      "step": 17358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7945777002719128,
+      "learning_rate": 4.292372789947155e-06,
+      "loss": 1.1875,
+      "step": 17359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0908140407790685,
+      "learning_rate": 4.2900364129465245e-06,
+      "loss": 1.2656,
+      "step": 17360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.5691888976728885,
+      "learning_rate": 4.287700612295742e-06,
+      "loss": 0.9922,
+      "step": 17361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4605904470686637,
+      "learning_rate": 4.285365388059826e-06,
+      "loss": 1.0957,
+      "step": 17362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0504457035759294,
+      "learning_rate": 4.283030740303756e-06,
+      "loss": 0.9731,
+      "step": 17363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5001925503805047,
+      "learning_rate": 4.2806966690925045e-06,
+      "loss": 0.3738,
+      "step": 17364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.231575445808838,
+      "learning_rate": 4.278363174491032e-06,
+      "loss": 0.8691,
+      "step": 17365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.00775822507396,
+      "learning_rate": 4.276030256564279e-06,
+      "loss": 0.8105,
+      "step": 17366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.398678815330186,
+      "learning_rate": 4.273697915377167e-06,
+      "loss": 1.166,
+      "step": 17367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6447261586339326,
+      "learning_rate": 4.271366150994599e-06,
+      "loss": 1.1133,
+      "step": 17368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8845242842936316,
+      "learning_rate": 4.269034963481477e-06,
+      "loss": 0.9004,
+      "step": 17369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7465509980885696,
+      "learning_rate": 4.266704352902668e-06,
+      "loss": 1.0166,
+      "step": 17370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.623942843506693,
+      "learning_rate": 4.264374319323039e-06,
+      "loss": 1.377,
+      "step": 17371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.590477798540428,
+      "learning_rate": 4.262044862807435e-06,
+      "loss": 1.2441,
+      "step": 17372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7811421577033744,
+      "learning_rate": 4.259715983420673e-06,
+      "loss": 1.2539,
+      "step": 17373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5127425650798885,
+      "learning_rate": 4.2573876812275755e-06,
+      "loss": 1.0918,
+      "step": 17374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7927633570128934,
+      "learning_rate": 4.255059956292936e-06,
+      "loss": 1.1006,
+      "step": 17375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.207434872903686,
+      "learning_rate": 4.252732808681528e-06,
+      "loss": 1.7275,
+      "step": 17376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.69500633380549,
+      "learning_rate": 4.250406238458124e-06,
+      "loss": 0.8525,
+      "step": 17377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.140738565849388,
+      "learning_rate": 4.248080245687469e-06,
+      "loss": 1.1055,
+      "step": 17378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.744045834244718,
+      "learning_rate": 4.24575483043429e-06,
+      "loss": 0.615,
+      "step": 17379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.635348485885593,
+      "learning_rate": 4.243429992763301e-06,
+      "loss": 1.1875,
+      "step": 17380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.45342325717833,
+      "learning_rate": 4.2411057327392065e-06,
+      "loss": 0.8164,
+      "step": 17381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.449545683870539,
+      "learning_rate": 4.23878205042669e-06,
+      "loss": 0.9736,
+      "step": 17382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.1410058526876,
+      "learning_rate": 4.23645894589042e-06,
+      "loss": 1.4941,
+      "step": 17383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4661421036598705,
+      "learning_rate": 4.234136419195045e-06,
+      "loss": 1.2773,
+      "step": 17384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.255138208316505,
+      "learning_rate": 4.2318144704051935e-06,
+      "loss": 1.0134,
+      "step": 17385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.568028484102018,
+      "learning_rate": 4.229493099585497e-06,
+      "loss": 1.2314,
+      "step": 17386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.812370052730883,
+      "learning_rate": 4.22717230680055e-06,
+      "loss": 1.1113,
+      "step": 17387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0784354300711625,
+      "learning_rate": 4.224852092114945e-06,
+      "loss": 0.755,
+      "step": 17388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5961010859635762,
+      "learning_rate": 4.222532455593239e-06,
+      "loss": 1.0283,
+      "step": 17389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.364235260264822,
+      "learning_rate": 4.220213397299996e-06,
+      "loss": 1.4844,
+      "step": 17390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4605854085305268,
+      "learning_rate": 4.217894917299764e-06,
+      "loss": 0.6055,
+      "step": 17391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8926427638315345,
+      "learning_rate": 4.2155770156570546e-06,
+      "loss": 0.0533,
+      "step": 17392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.400273110135529,
+      "learning_rate": 4.213259692436367e-06,
+      "loss": 0.8867,
+      "step": 17393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.036992935899723,
+      "learning_rate": 4.21094294770221e-06,
+      "loss": 1.1313,
+      "step": 17394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.421171203139144,
+      "learning_rate": 4.208626781519043e-06,
+      "loss": 0.8994,
+      "step": 17395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.242425567004057,
+      "learning_rate": 4.206311193951332e-06,
+      "loss": 0.8398,
+      "step": 17396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.528891983911192,
+      "learning_rate": 4.203996185063508e-06,
+      "loss": 1.4512,
+      "step": 17397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.83682607393067,
+      "learning_rate": 4.201681754920004e-06,
+      "loss": 1.3125,
+      "step": 17398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2615076997681967,
+      "learning_rate": 4.1993679035852355e-06,
+      "loss": 0.7158,
+      "step": 17399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1189906514802184,
+      "learning_rate": 4.1970546311235926e-06,
+      "loss": 0.9453,
+      "step": 17400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.37835606819662,
+      "learning_rate": 4.1947419375994476e-06,
+      "loss": 1.6992,
+      "step": 17401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.622707001232266,
+      "learning_rate": 4.19242982307716e-06,
+      "loss": 1.3672,
+      "step": 17402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.98912413434356,
+      "learning_rate": 4.190118287621087e-06,
+      "loss": 1.3086,
+      "step": 17403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7289649850032482,
+      "learning_rate": 4.187807331295549e-06,
+      "loss": 1.0635,
+      "step": 17404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.640153073764006,
+      "learning_rate": 4.185496954164858e-06,
+      "loss": 2.2109,
+      "step": 17405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6967801286226116,
+      "learning_rate": 4.183187156293311e-06,
+      "loss": 0.9814,
+      "step": 17406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5400631734092705,
+      "learning_rate": 4.1808779377451865e-06,
+      "loss": 1.4736,
+      "step": 17407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.10925478451181,
+      "learning_rate": 4.178569298584764e-06,
+      "loss": 1.0344,
+      "step": 17408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.585156723491266,
+      "learning_rate": 4.1762612388762805e-06,
+      "loss": 1.3467,
+      "step": 17409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7749364965922427,
+      "learning_rate": 4.1739537586839685e-06,
+      "loss": 1.2617,
+      "step": 17410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.717582802866509,
+      "learning_rate": 4.17164685807204e-06,
+      "loss": 0.772,
+      "step": 17411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0737382292686495,
+      "learning_rate": 4.169340537104704e-06,
+      "loss": 0.9512,
+      "step": 17412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.966778552342203,
+      "learning_rate": 4.167034795846142e-06,
+      "loss": 0.8857,
+      "step": 17413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.1143621203166,
+      "learning_rate": 4.164729634360515e-06,
+      "loss": 1.2344,
+      "step": 17414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5220868065435833,
+      "learning_rate": 4.162425052711985e-06,
+      "loss": 0.9424,
+      "step": 17415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.345269994585626,
+      "learning_rate": 4.160121050964685e-06,
+      "loss": 1.2627,
+      "step": 17416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.122136936116202,
+      "learning_rate": 4.157817629182725e-06,
+      "loss": 0.6448,
+      "step": 17417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3806527778044444,
+      "learning_rate": 4.1555147874302206e-06,
+      "loss": 0.8403,
+      "step": 17418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.650348885080318,
+      "learning_rate": 4.15321252577125e-06,
+      "loss": 1.1152,
+      "step": 17419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9800593989546555,
+      "learning_rate": 4.150910844269892e-06,
+      "loss": 1.2588,
+      "step": 17420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6334787954279735,
+      "learning_rate": 4.148609742990198e-06,
+      "loss": 0.7275,
+      "step": 17421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.450111627785966,
+      "learning_rate": 4.1463092219962055e-06,
+      "loss": 0.4417,
+      "step": 17422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.254674083223031,
+      "learning_rate": 4.144009281351935e-06,
+      "loss": 0.5327,
+      "step": 17423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.159627359646314,
+      "learning_rate": 4.141709921121398e-06,
+      "loss": 1.0723,
+      "step": 17424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5101980111488778,
+      "learning_rate": 4.139411141368585e-06,
+      "loss": 1.2246,
+      "step": 17425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5621429752892353,
+      "learning_rate": 4.137112942157462e-06,
+      "loss": 0.9961,
+      "step": 17426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.702002504089267,
+      "learning_rate": 4.1348153235519965e-06,
+      "loss": 1.8398,
+      "step": 17427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7917099733071997,
+      "learning_rate": 4.132518285616122e-06,
+      "loss": 1.0703,
+      "step": 17428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9578410470012308,
+      "learning_rate": 4.130221828413772e-06,
+      "loss": 0.9932,
+      "step": 17429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.40018352363169,
+      "learning_rate": 4.127925952008854e-06,
+      "loss": 0.833,
+      "step": 17430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2850616174649585,
+      "learning_rate": 4.125630656465259e-06,
+      "loss": 0.792,
+      "step": 17431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.163613613685154,
+      "learning_rate": 4.123335941846859e-06,
+      "loss": 1.0166,
+      "step": 17432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.043651011652546,
+      "learning_rate": 4.121041808217529e-06,
+      "loss": 1.3418,
+      "step": 17433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7440680316221604,
+      "learning_rate": 4.118748255641103e-06,
+      "loss": 1.3682,
+      "step": 17434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6023226707060565,
+      "learning_rate": 4.1164552841814085e-06,
+      "loss": 1.1406,
+      "step": 17435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1075434058572875,
+      "learning_rate": 4.11416289390226e-06,
+      "loss": 1.3867,
+      "step": 17436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6138088908190302,
+      "learning_rate": 4.111871084867461e-06,
+      "loss": 0.834,
+      "step": 17437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.192140821422361,
+      "learning_rate": 4.109579857140788e-06,
+      "loss": 1.2114,
+      "step": 17438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.107534048647643,
+      "learning_rate": 4.107289210786003e-06,
+      "loss": 1.5273,
+      "step": 17439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.269799996911296,
+      "learning_rate": 4.104999145866847e-06,
+      "loss": 0.9971,
+      "step": 17440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4510933791278426,
+      "learning_rate": 4.102709662447066e-06,
+      "loss": 0.9722,
+      "step": 17441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5588226077273766,
+      "learning_rate": 4.100420760590365e-06,
+      "loss": 1.4316,
+      "step": 17442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.273890274303053,
+      "learning_rate": 4.098132440360447e-06,
+      "loss": 0.8135,
+      "step": 17443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.600785147886601,
+      "learning_rate": 4.095844701820989e-06,
+      "loss": 0.9912,
+      "step": 17444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5950670970807006,
+      "learning_rate": 4.0935575450356615e-06,
+      "loss": 0.9719,
+      "step": 17445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0541596798125146,
+      "learning_rate": 4.091270970068123e-06,
+      "loss": 1.1436,
+      "step": 17446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1365513709864916,
+      "learning_rate": 4.0889849769820024e-06,
+      "loss": 1.0566,
+      "step": 17447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.066528006888052,
+      "learning_rate": 4.086699565840915e-06,
+      "loss": 0.6885,
+      "step": 17448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.45864017928031,
+      "learning_rate": 4.084414736708461e-06,
+      "loss": 1.8242,
+      "step": 17449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.044800506972222,
+      "learning_rate": 4.082130489648236e-06,
+      "loss": 1.1797,
+      "step": 17450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0279577718246995,
+      "learning_rate": 4.079846824723804e-06,
+      "loss": 0.7095,
+      "step": 17451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8311143114193977,
+      "learning_rate": 4.0775637419987164e-06,
+      "loss": 0.7808,
+      "step": 17452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0414607534567137,
+      "learning_rate": 4.075281241536508e-06,
+      "loss": 0.9702,
+      "step": 17453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9948888676984398,
+      "learning_rate": 4.072999323400703e-06,
+      "loss": 0.8687,
+      "step": 17454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6835458631837823,
+      "learning_rate": 4.070717987654815e-06,
+      "loss": 0.7407,
+      "step": 17455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0219162852357684,
+      "learning_rate": 4.068437234362324e-06,
+      "loss": 0.9912,
+      "step": 17456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.42300359319462,
+      "learning_rate": 4.066157063586698e-06,
+      "loss": 1.166,
+      "step": 17457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.530617505358227,
+      "learning_rate": 4.063877475391404e-06,
+      "loss": 1.3984,
+      "step": 17458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3765280761093757,
+      "learning_rate": 4.061598469839878e-06,
+      "loss": 1.3047,
+      "step": 17459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7360144762059675,
+      "learning_rate": 4.05932004699554e-06,
+      "loss": 0.8491,
+      "step": 17460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.252017494201467,
+      "learning_rate": 4.057042206921796e-06,
+      "loss": 0.4473,
+      "step": 17461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.417103472358852,
+      "learning_rate": 4.0547649496820465e-06,
+      "loss": 0.9424,
+      "step": 17462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2884365763812937,
+      "learning_rate": 4.052488275339658e-06,
+      "loss": 1.3281,
+      "step": 17463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2414562348742044,
+      "learning_rate": 4.050212183957996e-06,
+      "loss": 1.3262,
+      "step": 17464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.478422780328498,
+      "learning_rate": 4.047936675600403e-06,
+      "loss": 1.4023,
+      "step": 17465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.976511877235209,
+      "learning_rate": 4.045661750330196e-06,
+      "loss": 1.1646,
+      "step": 17466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1242290308736194,
+      "learning_rate": 4.043387408210697e-06,
+      "loss": 1.0527,
+      "step": 17467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9220450922947907,
+      "learning_rate": 4.0411136493051944e-06,
+      "loss": 0.8172,
+      "step": 17468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.185823724715647,
+      "learning_rate": 4.038840473676969e-06,
+      "loss": 0.8682,
+      "step": 17469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0530775601786533,
+      "learning_rate": 4.036567881389272e-06,
+      "loss": 1.0469,
+      "step": 17470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7345700540919327,
+      "learning_rate": 4.034295872505364e-06,
+      "loss": 0.7318,
+      "step": 17471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8575588385494646,
+      "learning_rate": 4.032024447088459e-06,
+      "loss": 0.9033,
+      "step": 17472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.788353427391129,
+      "learning_rate": 4.029753605201786e-06,
+      "loss": 0.6597,
+      "step": 17473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7635961708558516,
+      "learning_rate": 4.027483346908525e-06,
+      "loss": 1.1396,
+      "step": 17474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.133859454288004,
+      "learning_rate": 4.025213672271872e-06,
+      "loss": 0.9703,
+      "step": 17475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.975075017159576,
+      "learning_rate": 4.022944581354981e-06,
+      "loss": 0.6052,
+      "step": 17476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0212777874086596,
+      "learning_rate": 4.020676074221005e-06,
+      "loss": 1.2832,
+      "step": 17477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.442240510567948,
+      "learning_rate": 4.018408150933067e-06,
+      "loss": 1.2246,
+      "step": 17478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4684074260745232,
+      "learning_rate": 4.0161408115542944e-06,
+      "loss": 1.8398,
+      "step": 17479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3404955846914053,
+      "learning_rate": 4.013874056147779e-06,
+      "loss": 0.9004,
+      "step": 17480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8166740808633888,
+      "learning_rate": 4.011607884776608e-06,
+      "loss": 1.1865,
+      "step": 17481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0683484966587877,
+      "learning_rate": 4.0093422975038375e-06,
+      "loss": 1.7109,
+      "step": 17482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8758705025776155,
+      "learning_rate": 4.007077294392525e-06,
+      "loss": 0.9106,
+      "step": 17483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.226333425301207,
+      "learning_rate": 4.004812875505712e-06,
+      "loss": 0.8916,
+      "step": 17484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.703268322976297,
+      "learning_rate": 4.0025490409064085e-06,
+      "loss": 0.7305,
+      "step": 17485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5825542574435436,
+      "learning_rate": 4.000285790657618e-06,
+      "loss": 0.8193,
+      "step": 17486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7566659808108884,
+      "learning_rate": 3.99802312482232e-06,
+      "loss": 1.457,
+      "step": 17487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7277286516916455,
+      "learning_rate": 3.995761043463495e-06,
+      "loss": 0.8164,
+      "step": 17488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.818876793963709,
+      "learning_rate": 3.993499546644086e-06,
+      "loss": 1.1631,
+      "step": 17489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.661376404418833,
+      "learning_rate": 3.991238634427036e-06,
+      "loss": 1.022,
+      "step": 17490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.2152590468412185,
+      "learning_rate": 3.988978306875249e-06,
+      "loss": 1.0967,
+      "step": 17491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.016503194483415,
+      "learning_rate": 3.986718564051653e-06,
+      "loss": 1.4453,
+      "step": 17492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0703871227707067,
+      "learning_rate": 3.984459406019125e-06,
+      "loss": 0.6025,
+      "step": 17493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.2144806074633,
+      "learning_rate": 3.982200832840538e-06,
+      "loss": 1.5234,
+      "step": 17494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0720049004662493,
+      "learning_rate": 3.979942844578738e-06,
+      "loss": 0.3137,
+      "step": 17495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.25924585220039,
+      "learning_rate": 3.977685441296578e-06,
+      "loss": 0.8311,
+      "step": 17496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9930119024171105,
+      "learning_rate": 3.9754286230568734e-06,
+      "loss": 0.6243,
+      "step": 17497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.713497851538702,
+      "learning_rate": 3.973172389922433e-06,
+      "loss": 0.9971,
+      "step": 17498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3038868812592845,
+      "learning_rate": 3.970916741956038e-06,
+      "loss": 1.3018,
+      "step": 17499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.138554123575606,
+      "learning_rate": 3.968661679220468e-06,
+      "loss": 1.127,
+      "step": 17500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.124491342336387,
+      "learning_rate": 3.9664072017784885e-06,
+      "loss": 0.9927,
+      "step": 17501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.213548455121184,
+      "learning_rate": 3.964153309692833e-06,
+      "loss": 0.9951,
+      "step": 17502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3409624943782097,
+      "learning_rate": 3.961900003026229e-06,
+      "loss": 1.1465,
+      "step": 17503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8853039490795824,
+      "learning_rate": 3.959647281841375e-06,
+      "loss": 0.9395,
+      "step": 17504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9916963458011272,
+      "learning_rate": 3.957395146200979e-06,
+      "loss": 0.7227,
+      "step": 17505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8092543382222348,
+      "learning_rate": 3.95514359616771e-06,
+      "loss": 1.2246,
+      "step": 17506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.297997647954071,
+      "learning_rate": 3.9528926318042255e-06,
+      "loss": 0.5472,
+      "step": 17507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.053069984312738,
+      "learning_rate": 3.950642253173167e-06,
+      "loss": 1.21,
+      "step": 17508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.919782082774102,
+      "learning_rate": 3.9483924603371635e-06,
+      "loss": 1.1582,
+      "step": 17509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2039235068061096,
+      "learning_rate": 3.946143253358836e-06,
+      "loss": 0.5989,
+      "step": 17510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.179168840118076,
+      "learning_rate": 3.943894632300768e-06,
+      "loss": 1.4727,
+      "step": 17511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.289087022192045,
+      "learning_rate": 3.941646597225537e-06,
+      "loss": 0.7411,
+      "step": 17512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1936680075913344,
+      "learning_rate": 3.9393991481957135e-06,
+      "loss": 0.8623,
+      "step": 17513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6195933042921498,
+      "learning_rate": 3.937152285273837e-06,
+      "loss": 1.0273,
+      "step": 17514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0755176635691397,
+      "learning_rate": 3.934906008522437e-06,
+      "loss": 0.9331,
+      "step": 17515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.960062539269106,
+      "learning_rate": 3.9326603180040225e-06,
+      "loss": 0.729,
+      "step": 17516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6308439689020893,
+      "learning_rate": 3.930415213781102e-06,
+      "loss": 1.416,
+      "step": 17517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9137765476171962,
+      "learning_rate": 3.9281706959161475e-06,
+      "loss": 1.4043,
+      "step": 17518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9978414450424786,
+      "learning_rate": 3.925926764471619e-06,
+      "loss": 0.8682,
+      "step": 17519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.668198046458417,
+      "learning_rate": 3.923683419509974e-06,
+      "loss": 0.7188,
+      "step": 17520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8191134515548457,
+      "learning_rate": 3.921440661093634e-06,
+      "loss": 1.7129,
+      "step": 17521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.7704016274976,
+      "learning_rate": 3.919198489285025e-06,
+      "loss": 0.8196,
+      "step": 17522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5389728810509395,
+      "learning_rate": 3.91695690414654e-06,
+      "loss": 0.4579,
+      "step": 17523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.671558956328619,
+      "learning_rate": 3.914715905740562e-06,
+      "loss": 0.7112,
+      "step": 17524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2319113213482082,
+      "learning_rate": 3.91247549412945e-06,
+      "loss": 0.8789,
+      "step": 17525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1681761573789693,
+      "learning_rate": 3.910235669375567e-06,
+      "loss": 1.2617,
+      "step": 17526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8377635951008475,
+      "learning_rate": 3.907996431541239e-06,
+      "loss": 0.873,
+      "step": 17527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.809281681797034,
+      "learning_rate": 3.905757780688776e-06,
+      "loss": 0.5117,
+      "step": 17528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.084607900933754,
+      "learning_rate": 3.903519716880494e-06,
+      "loss": 1.1758,
+      "step": 17529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9262028429874962,
+      "learning_rate": 3.9012822401786665e-06,
+      "loss": 1.3516,
+      "step": 17530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2232406548710983,
+      "learning_rate": 3.899045350645567e-06,
+      "loss": 1.2051,
+      "step": 17531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.121340011997767,
+      "learning_rate": 3.896809048343447e-06,
+      "loss": 1.2656,
+      "step": 17532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8771500657808526,
+      "learning_rate": 3.894573333334534e-06,
+      "loss": 0.9238,
+      "step": 17533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8812171223451106,
+      "learning_rate": 3.892338205681059e-06,
+      "loss": 0.9131,
+      "step": 17534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0420030033415264,
+      "learning_rate": 3.890103665445219e-06,
+      "loss": 1.2246,
+      "step": 17535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.508730236178992,
+      "learning_rate": 3.887869712689199e-06,
+      "loss": 1.1162,
+      "step": 17536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5237979218164033,
+      "learning_rate": 3.885636347475166e-06,
+      "loss": 0.8066,
+      "step": 17537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3937916445782386,
+      "learning_rate": 3.883403569865277e-06,
+      "loss": 0.5491,
+      "step": 17538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0518772647878865,
+      "learning_rate": 3.8811713799216776e-06,
+      "loss": 1.5449,
+      "step": 17539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4218011660242356,
+      "learning_rate": 3.878939777706478e-06,
+      "loss": 0.9102,
+      "step": 17540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.2551229679256,
+      "learning_rate": 3.876708763281789e-06,
+      "loss": 0.9854,
+      "step": 17541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.971411225789721,
+      "learning_rate": 3.874478336709686e-06,
+      "loss": 1.5527,
+      "step": 17542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.071692392038329,
+      "learning_rate": 3.872248498052261e-06,
+      "loss": 0.3882,
+      "step": 17543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2959803230568636,
+      "learning_rate": 3.870019247371556e-06,
+      "loss": 0.8584,
+      "step": 17544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0886930060472606,
+      "learning_rate": 3.867790584729617e-06,
+      "loss": 0.9785,
+      "step": 17545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.317287080970787,
+      "learning_rate": 3.865562510188453e-06,
+      "loss": 1.5996,
+      "step": 17546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.212950182440096,
+      "learning_rate": 3.863335023810083e-06,
+      "loss": 1.6064,
+      "step": 17547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.209549132677486,
+      "learning_rate": 3.861108125656499e-06,
+      "loss": 0.7144,
+      "step": 17548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.713240280491797,
+      "learning_rate": 3.858881815789672e-06,
+      "loss": 1.7051,
+      "step": 17549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7072902627089763,
+      "learning_rate": 3.856656094271549e-06,
+      "loss": 0.7588,
+      "step": 17550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4546992653275885,
+      "learning_rate": 3.854430961164088e-06,
+      "loss": 0.9814,
+      "step": 17551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5763397509152393,
+      "learning_rate": 3.852206416529203e-06,
+      "loss": 0.9375,
+      "step": 17552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.313873514569827,
+      "learning_rate": 3.849982460428805e-06,
+      "loss": 0.9834,
+      "step": 17553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.452451674270082,
+      "learning_rate": 3.84775909292478e-06,
+      "loss": 1.2773,
+      "step": 17554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.607768704564466,
+      "learning_rate": 3.845536314079012e-06,
+      "loss": 1.1074,
+      "step": 17555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.305900022008548,
+      "learning_rate": 3.843314123953354e-06,
+      "loss": 1.0518,
+      "step": 17556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0577833400968806,
+      "learning_rate": 3.841092522609655e-06,
+      "loss": 1.0859,
+      "step": 17557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.291770296932611,
+      "learning_rate": 3.838871510109737e-06,
+      "loss": 1.3496,
+      "step": 17558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0450137982737036,
+      "learning_rate": 3.8366510865154025e-06,
+      "loss": 1.0635,
+      "step": 17559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9955437008730519,
+      "learning_rate": 3.834431251888462e-06,
+      "loss": 1.2705,
+      "step": 17560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.1112376151814765,
+      "learning_rate": 3.832212006290681e-06,
+      "loss": 1.2871,
+      "step": 17561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.815077384847213,
+      "learning_rate": 3.829993349783822e-06,
+      "loss": 1.0586,
+      "step": 17562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.828705617410674,
+      "learning_rate": 3.827775282429627e-06,
+      "loss": 0.7354,
+      "step": 17563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.454435127770943,
+      "learning_rate": 3.825557804289831e-06,
+      "loss": 1.2588,
+      "step": 17564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0199290182601297,
+      "learning_rate": 3.823340915426135e-06,
+      "loss": 1.0498,
+      "step": 17565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.370275507781634,
+      "learning_rate": 3.821124615900246e-06,
+      "loss": 1.1641,
+      "step": 17566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1474866922232043,
+      "learning_rate": 3.818908905773835e-06,
+      "loss": 0.4137,
+      "step": 17567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3983715533292616,
+      "learning_rate": 3.816693785108563e-06,
+      "loss": 1.0781,
+      "step": 17568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.114310677506275,
+      "learning_rate": 3.8144792539660856e-06,
+      "loss": 1.2441,
+      "step": 17569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9068253470629495,
+      "learning_rate": 3.8122653124080254e-06,
+      "loss": 0.6621,
+      "step": 17570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2322532586780484,
+      "learning_rate": 3.810051960495989e-06,
+      "loss": 0.9482,
+      "step": 17571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.714845368164594,
+      "learning_rate": 3.807839198291588e-06,
+      "loss": 0.7006,
+      "step": 17572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.413745649463243,
+      "learning_rate": 3.8056270258563924e-06,
+      "loss": 0.7251,
+      "step": 17573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1059886736298066,
+      "learning_rate": 3.803415443251965e-06,
+      "loss": 0.9199,
+      "step": 17574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5422880840977022,
+      "learning_rate": 3.8012044505398607e-06,
+      "loss": 1.1025,
+      "step": 17575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0566595009900395,
+      "learning_rate": 3.798994047781601e-06,
+      "loss": 1.1367,
+      "step": 17576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.584731175004433,
+      "learning_rate": 3.796784235038711e-06,
+      "loss": 1.3193,
+      "step": 17577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7583871734374026,
+      "learning_rate": 3.7945750123726865e-06,
+      "loss": 1.2773,
+      "step": 17578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3467132146871514,
+      "learning_rate": 3.7923663798450056e-06,
+      "loss": 0.9453,
+      "step": 17579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4186876550175263,
+      "learning_rate": 3.7901583375171273e-06,
+      "loss": 1.0488,
+      "step": 17580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.034528394882095,
+      "learning_rate": 3.787950885450514e-06,
+      "loss": 1.2383,
+      "step": 17581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.798815949329781,
+      "learning_rate": 3.7857440237065946e-06,
+      "loss": 1.3809,
+      "step": 17582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9354084766527277,
+      "learning_rate": 3.7835377523467744e-06,
+      "loss": 1.0996,
+      "step": 17583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9905220356818036,
+      "learning_rate": 3.7813320714324697e-06,
+      "loss": 0.9685,
+      "step": 17584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.118212809284351,
+      "learning_rate": 3.7791269810250467e-06,
+      "loss": 0.8672,
+      "step": 17585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.910603351963281,
+      "learning_rate": 3.7769224811858876e-06,
+      "loss": 0.9609,
+      "step": 17586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5049738283581497,
+      "learning_rate": 3.7747185719763352e-06,
+      "loss": 0.7581,
+      "step": 17587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9653753998514656,
+      "learning_rate": 3.7725152534577256e-06,
+      "loss": 0.7352,
+      "step": 17588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.402955611550397,
+      "learning_rate": 3.770312525691369e-06,
+      "loss": 0.9668,
+      "step": 17589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6153105880651917,
+      "learning_rate": 3.768110388738577e-06,
+      "loss": 0.713,
+      "step": 17590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.4595763473125745,
+      "learning_rate": 3.765908842660629e-06,
+      "loss": 0.751,
+      "step": 17591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1423302979179466,
+      "learning_rate": 3.7637078875187886e-06,
+      "loss": 0.687,
+      "step": 17592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.566978161906364,
+      "learning_rate": 3.7615075233743134e-06,
+      "loss": 1.2188,
+      "step": 17593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7996370292779615,
+      "learning_rate": 3.759307750288443e-06,
+      "loss": 1.6543,
+      "step": 17594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.84195090076207,
+      "learning_rate": 3.75710856832239e-06,
+      "loss": 0.9141,
+      "step": 17595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.606374046268174,
+      "learning_rate": 3.7549099775373576e-06,
+      "loss": 0.6621,
+      "step": 17596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6245265715422743,
+      "learning_rate": 3.7527119779945245e-06,
+      "loss": 1.2441,
+      "step": 17597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.631226919621161,
+      "learning_rate": 3.750514569755076e-06,
+      "loss": 0.7019,
+      "step": 17598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.0593543002644115,
+      "learning_rate": 3.748317752880154e-06,
+      "loss": 1.418,
+      "step": 17599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.381467045381591,
+      "learning_rate": 3.7461215274308986e-06,
+      "loss": 1.0713,
+      "step": 17600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5140266790499366,
+      "learning_rate": 3.7439258934684224e-06,
+      "loss": 1.2422,
+      "step": 17601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9979994779437464,
+      "learning_rate": 3.7417308510538348e-06,
+      "loss": 1.5312,
+      "step": 17602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.141877988851933,
+      "learning_rate": 3.739536400248231e-06,
+      "loss": 0.8823,
+      "step": 17603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1044727014759124,
+      "learning_rate": 3.7373425411126705e-06,
+      "loss": 0.7734,
+      "step": 17604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0609956514774446,
+      "learning_rate": 3.7351492737082135e-06,
+      "loss": 1.1191,
+      "step": 17605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4516529112775096,
+      "learning_rate": 3.732956598095891e-06,
+      "loss": 0.9746,
+      "step": 17606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.84764824960279,
+      "learning_rate": 3.730764514336732e-06,
+      "loss": 1.0913,
+      "step": 17607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.806909313073838,
+      "learning_rate": 3.728573022491738e-06,
+      "loss": 1.1914,
+      "step": 17608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.21869645690862,
+      "learning_rate": 3.7263821226218994e-06,
+      "loss": 1.5879,
+      "step": 17609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4938421705608156,
+      "learning_rate": 3.7241918147881777e-06,
+      "loss": 0.9355,
+      "step": 17610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3455618302024694,
+      "learning_rate": 3.722002099051536e-06,
+      "loss": 1.0762,
+      "step": 17611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.740785796946977,
+      "learning_rate": 3.7198129754729194e-06,
+      "loss": 0.9744,
+      "step": 17612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.807361224005848,
+      "learning_rate": 3.717624444113246e-06,
+      "loss": 0.707,
+      "step": 17613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.165515948655899,
+      "learning_rate": 3.715436505033415e-06,
+      "loss": 1.3965,
+      "step": 17614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3754074415881288,
+      "learning_rate": 3.7132491582943267e-06,
+      "loss": 0.4131,
+      "step": 17615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8027623617124036,
+      "learning_rate": 3.711062403956847e-06,
+      "loss": 1.168,
+      "step": 17616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.204621329301345,
+      "learning_rate": 3.708876242081835e-06,
+      "loss": 1.5273,
+      "step": 17617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5596085959973194,
+      "learning_rate": 3.706690672730123e-06,
+      "loss": 1.2979,
+      "step": 17618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3014158101674402,
+      "learning_rate": 3.7045056959625457e-06,
+      "loss": 1.0181,
+      "step": 17619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.049548500952987,
+      "learning_rate": 3.702321311839907e-06,
+      "loss": 0.7844,
+      "step": 17620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9129549841710785,
+      "learning_rate": 3.7001375204229914e-06,
+      "loss": 1.1338,
+      "step": 17621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.145519030366523,
+      "learning_rate": 3.6979543217725815e-06,
+      "loss": 1.1465,
+      "step": 17622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.749721252879349,
+      "learning_rate": 3.6957717159494255e-06,
+      "loss": 1.166,
+      "step": 17623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.639563623689461,
+      "learning_rate": 3.6935897030142747e-06,
+      "loss": 0.6484,
+      "step": 17624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.154449485152977,
+      "learning_rate": 3.69140828302785e-06,
+      "loss": 1.6934,
+      "step": 17625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9670249405187838,
+      "learning_rate": 3.689227456050856e-06,
+      "loss": 1.2148,
+      "step": 17626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.149052202544809,
+      "learning_rate": 3.687047222143983e-06,
+      "loss": 0.9717,
+      "step": 17627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.538426802599625,
+      "learning_rate": 3.6848675813679152e-06,
+      "loss": 1.4336,
+      "step": 17628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.552122427239543,
+      "learning_rate": 3.682688533783302e-06,
+      "loss": 0.9365,
+      "step": 17629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.670981700264489,
+      "learning_rate": 3.680510079450786e-06,
+      "loss": 0.6758,
+      "step": 17630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.0258391622962595,
+      "learning_rate": 3.678332218430994e-06,
+      "loss": 0.6943,
+      "step": 17631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.436999551523859,
+      "learning_rate": 3.676154950784541e-06,
+      "loss": 1.543,
+      "step": 17632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.49539093809932,
+      "learning_rate": 3.6739782765720156e-06,
+      "loss": 1.3965,
+      "step": 17633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.168599025077569,
+      "learning_rate": 3.671802195853993e-06,
+      "loss": 1.3086,
+      "step": 17634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5960107364436826,
+      "learning_rate": 3.6696267086910293e-06,
+      "loss": 0.6062,
+      "step": 17635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.15255366429074,
+      "learning_rate": 3.6674518151436744e-06,
+      "loss": 1.2881,
+      "step": 17636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.658328869695314,
+      "learning_rate": 3.66527751527245e-06,
+      "loss": 0.9678,
+      "step": 17637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.205011770432115,
+      "learning_rate": 3.663103809137869e-06,
+      "loss": 1.4043,
+      "step": 17638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.494206869011876,
+      "learning_rate": 3.660930696800416e-06,
+      "loss": 1.6406,
+      "step": 17639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.143744267290149,
+      "learning_rate": 3.6587581783205733e-06,
+      "loss": 1.0557,
+      "step": 17640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0382417510516215,
+      "learning_rate": 3.6565862537588092e-06,
+      "loss": 1.1289,
+      "step": 17641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.38213541385536,
+      "learning_rate": 3.6544149231755613e-06,
+      "loss": 0.5748,
+      "step": 17642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.892602879034612,
+      "learning_rate": 3.652244186631254e-06,
+      "loss": 1.0474,
+      "step": 17643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.945248089652621,
+      "learning_rate": 3.6500740441862967e-06,
+      "loss": 1.4004,
+      "step": 17644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5045544146937813,
+      "learning_rate": 3.6479044959010946e-06,
+      "loss": 1.1855,
+      "step": 17645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.289480640974137,
+      "learning_rate": 3.645735541836015e-06,
+      "loss": 1.1221,
+      "step": 17646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.149583934372247,
+      "learning_rate": 3.6435671820514245e-06,
+      "loss": 0.9307,
+      "step": 17647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.052723108727111,
+      "learning_rate": 3.641399416607649e-06,
+      "loss": 0.6885,
+      "step": 17648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.72334094858525,
+      "learning_rate": 3.6392322455650495e-06,
+      "loss": 0.771,
+      "step": 17649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.625528918067629,
+      "learning_rate": 3.637065668983919e-06,
+      "loss": 1.2285,
+      "step": 17650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.080364211786747,
+      "learning_rate": 3.6348996869245535e-06,
+      "loss": 1.2222,
+      "step": 17651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.115996754382408,
+      "learning_rate": 3.632734299447224e-06,
+      "loss": 0.9619,
+      "step": 17652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2295701443129943,
+      "learning_rate": 3.630569506612208e-06,
+      "loss": 0.7695,
+      "step": 17653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.423501980741655,
+      "learning_rate": 3.628405308479743e-06,
+      "loss": 0.4766,
+      "step": 17654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8828219281794505,
+      "learning_rate": 3.626241705110059e-06,
+      "loss": 1.0869,
+      "step": 17655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.442367892471922,
+      "learning_rate": 3.62407869656336e-06,
+      "loss": 0.9961,
+      "step": 17656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.01016136777021,
+      "learning_rate": 3.6219162828998564e-06,
+      "loss": 0.7783,
+      "step": 17657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9939372438141796,
+      "learning_rate": 3.6197544641797116e-06,
+      "loss": 0.7166,
+      "step": 17658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.127264387126244,
+      "learning_rate": 3.6175932404631046e-06,
+      "loss": 1.0771,
+      "step": 17659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6445818266547008,
+      "learning_rate": 3.6154326118101713e-06,
+      "loss": 1.0479,
+      "step": 17660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.051282931802167,
+      "learning_rate": 3.613272578281035e-06,
+      "loss": 1.2559,
+      "step": 17661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9832897509866996,
+      "learning_rate": 3.611113139935826e-06,
+      "loss": 1.2451,
+      "step": 17662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.864028517827708,
+      "learning_rate": 3.6089542968346268e-06,
+      "loss": 1.5786,
+      "step": 17663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.988341448755637,
+      "learning_rate": 3.606796049037525e-06,
+      "loss": 1.04,
+      "step": 17664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2848669552968905,
+      "learning_rate": 3.6046383966045703e-06,
+      "loss": 0.8423,
+      "step": 17665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9307534057753113,
+      "learning_rate": 3.6024813395958284e-06,
+      "loss": 0.7197,
+      "step": 17666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1593896339145506,
+      "learning_rate": 3.6003248780713124e-06,
+      "loss": 1.1084,
+      "step": 17667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.672600986865852,
+      "learning_rate": 3.598169012091049e-06,
+      "loss": 0.8838,
+      "step": 17668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.247229071934025,
+      "learning_rate": 3.5960137417150234e-06,
+      "loss": 1.2012,
+      "step": 17669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.209460847088252,
+      "learning_rate": 3.59385906700323e-06,
+      "loss": 1.1504,
+      "step": 17670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2233869427571307,
+      "learning_rate": 3.5917049880156207e-06,
+      "loss": 0.6553,
+      "step": 17671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5977723518755815,
+      "learning_rate": 3.589551504812147e-06,
+      "loss": 1.4609,
+      "step": 17672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8700025495989703,
+      "learning_rate": 3.5873986174527363e-06,
+      "loss": 1.1777,
+      "step": 17673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4903028291755516,
+      "learning_rate": 3.5852463259973074e-06,
+      "loss": 1.0996,
+      "step": 17674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.557510449489751,
+      "learning_rate": 3.583094630505757e-06,
+      "loss": 1.0859,
+      "step": 17675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.126557844999153,
+      "learning_rate": 3.580943531037956e-06,
+      "loss": 1.0996,
+      "step": 17676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1935733455168682,
+      "learning_rate": 3.5787930276537847e-06,
+      "loss": 0.938,
+      "step": 17677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5334918608338044,
+      "learning_rate": 3.576643120413076e-06,
+      "loss": 0.8076,
+      "step": 17678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.162736921469835,
+      "learning_rate": 3.5744938093756758e-06,
+      "loss": 1.3809,
+      "step": 17679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6363104274498173,
+      "learning_rate": 3.572345094601387e-06,
+      "loss": 0.8994,
+      "step": 17680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8822297266733352,
+      "learning_rate": 3.5701969761500144e-06,
+      "loss": 0.9053,
+      "step": 17681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.004542632353405,
+      "learning_rate": 3.568049454081329e-06,
+      "loss": 0.5361,
+      "step": 17682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.963478410870814,
+      "learning_rate": 3.565902528455106e-06,
+      "loss": 1.4316,
+      "step": 17683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.211081309508057,
+      "learning_rate": 3.5637561993310917e-06,
+      "loss": 1.1304,
+      "step": 17684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2490023035268463,
+      "learning_rate": 3.561610466769011e-06,
+      "loss": 0.814,
+      "step": 17685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.580857741729372,
+      "learning_rate": 3.5594653308285873e-06,
+      "loss": 0.4395,
+      "step": 17686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.879387700106892,
+      "learning_rate": 3.55732079156951e-06,
+      "loss": 1.0264,
+      "step": 17687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.073830400430254,
+      "learning_rate": 3.5551768490514726e-06,
+      "loss": 1.2637,
+      "step": 17688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.085973617869661,
+      "learning_rate": 3.55303350333413e-06,
+      "loss": 1.0339,
+      "step": 17689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.782208116062289,
+      "learning_rate": 3.550890754477132e-06,
+      "loss": 0.8096,
+      "step": 17690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.573162422730876,
+      "learning_rate": 3.548748602540117e-06,
+      "loss": 1.1025,
+      "step": 17691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.24706394637299,
+      "learning_rate": 3.5466070475826925e-06,
+      "loss": 0.8628,
+      "step": 17692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6376030237406742,
+      "learning_rate": 3.5444660896644637e-06,
+      "loss": 0.7207,
+      "step": 17693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.300647569798008,
+      "learning_rate": 3.5423257288450027e-06,
+      "loss": 1.3418,
+      "step": 17694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.229132047077948,
+      "learning_rate": 3.5401859651838813e-06,
+      "loss": 0.7803,
+      "step": 17695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.444549683703014,
+      "learning_rate": 3.538046798740652e-06,
+      "loss": 0.5781,
+      "step": 17696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.133573292275763,
+      "learning_rate": 3.535908229574847e-06,
+      "loss": 1.0703,
+      "step": 17697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.828057237920676,
+      "learning_rate": 3.5337702577459747e-06,
+      "loss": 0.7422,
+      "step": 17698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5260139738467013,
+      "learning_rate": 3.531632883313532e-06,
+      "loss": 0.8701,
+      "step": 17699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0281459860164115,
+      "learning_rate": 3.529496106337013e-06,
+      "loss": 0.833,
+      "step": 17700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.033061636469823,
+      "learning_rate": 3.5273599268758757e-06,
+      "loss": 1.0078,
+      "step": 17701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0153506455611487,
+      "learning_rate": 3.5252243449895695e-06,
+      "loss": 1.3301,
+      "step": 17702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.252864391900284,
+      "learning_rate": 3.523089360737525e-06,
+      "loss": 0.7539,
+      "step": 17703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6634657644320257,
+      "learning_rate": 3.5209549741791582e-06,
+      "loss": 0.5349,
+      "step": 17704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.647575215540087,
+      "learning_rate": 3.5188211853738746e-06,
+      "loss": 0.8535,
+      "step": 17705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.589445004288394,
+      "learning_rate": 3.5166879943810543e-06,
+      "loss": 0.8291,
+      "step": 17706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.970279339841533,
+      "learning_rate": 3.5145554012600557e-06,
+      "loss": 0.748,
+      "step": 17707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.546703400104174,
+      "learning_rate": 3.5124234060702364e-06,
+      "loss": 0.7983,
+      "step": 17708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4741475493656258,
+      "learning_rate": 3.51029200887093e-06,
+      "loss": 1.1973,
+      "step": 17709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.601817167263852,
+      "learning_rate": 3.508161209721447e-06,
+      "loss": 1.3008,
+      "step": 17710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.392250198916398,
+      "learning_rate": 3.506031008681082e-06,
+      "loss": 0.5359,
+      "step": 17711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.243701383866262,
+      "learning_rate": 3.503901405809129e-06,
+      "loss": 0.7183,
+      "step": 17712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.126989265533263,
+      "learning_rate": 3.5017724011648434e-06,
+      "loss": 1.25,
+      "step": 17713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2548027701728484,
+      "learning_rate": 3.499643994807486e-06,
+      "loss": 1.1348,
+      "step": 17714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.010013444454648,
+      "learning_rate": 3.4975161867962847e-06,
+      "loss": 1.1328,
+      "step": 17715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.5683405868841955,
+      "learning_rate": 3.4953889771904475e-06,
+      "loss": 0.9336,
+      "step": 17716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6932548297954564,
+      "learning_rate": 3.4932623660491852e-06,
+      "loss": 0.542,
+      "step": 17717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4565695130156264,
+      "learning_rate": 3.491136353431679e-06,
+      "loss": 1.0117,
+      "step": 17718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.763289850473434,
+      "learning_rate": 3.489010939397089e-06,
+      "loss": 0.6895,
+      "step": 17719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.616585798868343,
+      "learning_rate": 3.4868861240045626e-06,
+      "loss": 1.3096,
+      "step": 17720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6493378861714474,
+      "learning_rate": 3.4847619073132413e-06,
+      "loss": 0.9219,
+      "step": 17721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9385587199214513,
+      "learning_rate": 3.4826382893822395e-06,
+      "loss": 1.25,
+      "step": 17722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6709452121132773,
+      "learning_rate": 3.4805152702706483e-06,
+      "loss": 1.127,
+      "step": 17723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.910027131154061,
+      "learning_rate": 3.4783928500375622e-06,
+      "loss": 1.252,
+      "step": 17724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.952679514756352,
+      "learning_rate": 3.476271028742037e-06,
+      "loss": 0.832,
+      "step": 17725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2935069815556655,
+      "learning_rate": 3.4741498064431303e-06,
+      "loss": 1.291,
+      "step": 17726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.050674338915765,
+      "learning_rate": 3.4720291831998736e-06,
+      "loss": 1.1055,
+      "step": 17727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.172062159096681,
+      "learning_rate": 3.4699091590712718e-06,
+      "loss": 1.2754,
+      "step": 17728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4065277137515877,
+      "learning_rate": 3.467789734116342e-06,
+      "loss": 1.3193,
+      "step": 17729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.76935089741699,
+      "learning_rate": 3.465670908394056e-06,
+      "loss": 0.8379,
+      "step": 17730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3916146376709526,
+      "learning_rate": 3.4635526819633835e-06,
+      "loss": 1.2207,
+      "step": 17731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.338144541406359,
+      "learning_rate": 3.4614350548832664e-06,
+      "loss": 0.832,
+      "step": 17732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.416758780210474,
+      "learning_rate": 3.4593180272126437e-06,
+      "loss": 0.9253,
+      "step": 17733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7650939007600606,
+      "learning_rate": 3.4572015990104347e-06,
+      "loss": 0.5195,
+      "step": 17734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.623497590051277,
+      "learning_rate": 3.4550857703355364e-06,
+      "loss": 0.8584,
+      "step": 17735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.9765420751284974,
+      "learning_rate": 3.4529705412468305e-06,
+      "loss": 0.5737,
+      "step": 17736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.572387850112397,
+      "learning_rate": 3.4508559118031775e-06,
+      "loss": 0.6111,
+      "step": 17737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.3652249807645,
+      "learning_rate": 3.4487418820634358e-06,
+      "loss": 1.3096,
+      "step": 17738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3036045439611583,
+      "learning_rate": 3.446628452086434e-06,
+      "loss": 0.9453,
+      "step": 17739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.607266516468297,
+      "learning_rate": 3.444515621930991e-06,
+      "loss": 0.6323,
+      "step": 17740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.936659729356204,
+      "learning_rate": 3.4424033916558936e-06,
+      "loss": 0.8784,
+      "step": 17741
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.880909980917625,
+      "learning_rate": 3.4402917613199358e-06,
+      "loss": 1.4043,
+      "step": 17742
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.155208612598409,
+      "learning_rate": 3.4381807309818854e-06,
+      "loss": 1.29,
+      "step": 17743
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.379286897331626,
+      "learning_rate": 3.436070300700489e-06,
+      "loss": 1.0342,
+      "step": 17744
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2920953754200486,
+      "learning_rate": 3.433960470534478e-06,
+      "loss": 1.2188,
+      "step": 17745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.34894564533747,
+      "learning_rate": 3.431851240542561e-06,
+      "loss": 1.0586,
+      "step": 17746
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6434179243043996,
+      "learning_rate": 3.4297426107834485e-06,
+      "loss": 1.2607,
+      "step": 17747
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1160311358638335,
+      "learning_rate": 3.4276345813158196e-06,
+      "loss": 1.0312,
+      "step": 17748
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.984071406555786,
+      "learning_rate": 3.4255271521983297e-06,
+      "loss": 0.665,
+      "step": 17749
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6756640128716744,
+      "learning_rate": 3.42342032348964e-06,
+      "loss": 0.6429,
+      "step": 17750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.387167297015846,
+      "learning_rate": 3.421314095248382e-06,
+      "loss": 1.1084,
+      "step": 17751
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.903830556881061,
+      "learning_rate": 3.4192084675331693e-06,
+      "loss": 0.7539,
+      "step": 17752
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.153294817362077,
+      "learning_rate": 3.417103440402597e-06,
+      "loss": 1.5977,
+      "step": 17753
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.556458745499456,
+      "learning_rate": 3.414999013915246e-06,
+      "loss": 1.2188,
+      "step": 17754
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6965354024896246,
+      "learning_rate": 3.4128951881296916e-06,
+      "loss": 1.6035,
+      "step": 17755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.262728187476974,
+      "learning_rate": 3.4107919631044732e-06,
+      "loss": 1.2891,
+      "step": 17756
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.469891923864295,
+      "learning_rate": 3.4086893388981274e-06,
+      "loss": 1.1777,
+      "step": 17757
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4158525686552945,
+      "learning_rate": 3.4065873155691625e-06,
+      "loss": 0.8167,
+      "step": 17758
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.021790754477753,
+      "learning_rate": 3.4044858931760847e-06,
+      "loss": 1.2363,
+      "step": 17759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.167762367191301,
+      "learning_rate": 3.40238507177737e-06,
+      "loss": 0.8721,
+      "step": 17760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.06668103953465,
+      "learning_rate": 3.40028485143149e-06,
+      "loss": 1.377,
+      "step": 17761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.0770360435481345,
+      "learning_rate": 3.3981852321968934e-06,
+      "loss": 0.9443,
+      "step": 17762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9282174062367807,
+      "learning_rate": 3.3960862141319967e-06,
+      "loss": 1.1445,
+      "step": 17763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.279904119184118,
+      "learning_rate": 3.393987797295234e-06,
+      "loss": 0.9312,
+      "step": 17764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1303229141820177,
+      "learning_rate": 3.3918899817449943e-06,
+      "loss": 1.0859,
+      "step": 17765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.450692009023084,
+      "learning_rate": 3.3897927675396623e-06,
+      "loss": 1.2324,
+      "step": 17766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9629688358626485,
+      "learning_rate": 3.387696154737591e-06,
+      "loss": 1.0874,
+      "step": 17767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7591748870821384,
+      "learning_rate": 3.385600143397144e-06,
+      "loss": 0.6843,
+      "step": 17768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6856153222811416,
+      "learning_rate": 3.3835047335766422e-06,
+      "loss": 0.7256,
+      "step": 17769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9001888156230478,
+      "learning_rate": 3.3814099253344084e-06,
+      "loss": 0.8438,
+      "step": 17770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9115286599384045,
+      "learning_rate": 3.3793157187287287e-06,
+      "loss": 1.0352,
+      "step": 17771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9088203213535495,
+      "learning_rate": 3.377222113817896e-06,
+      "loss": 1.1934,
+      "step": 17772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.336198631554693,
+      "learning_rate": 3.375129110660169e-06,
+      "loss": 1.7217,
+      "step": 17773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3720878013069786,
+      "learning_rate": 3.3730367093137984e-06,
+      "loss": 0.9297,
+      "step": 17774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5557642542812733,
+      "learning_rate": 3.370944909837004e-06,
+      "loss": 1.1582,
+      "step": 17775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.448034556911625,
+      "learning_rate": 3.3688537122880114e-06,
+      "loss": 0.5996,
+      "step": 17776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8359590219569024,
+      "learning_rate": 3.3667631167250163e-06,
+      "loss": 0.8301,
+      "step": 17777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3480049227697717,
+      "learning_rate": 3.3646731232061913e-06,
+      "loss": 1.0508,
+      "step": 17778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1632934746305574,
+      "learning_rate": 3.362583731789712e-06,
+      "loss": 1.3711,
+      "step": 17779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7566488525795134,
+      "learning_rate": 3.3604949425337094e-06,
+      "loss": 1.2852,
+      "step": 17780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8076681648209263,
+      "learning_rate": 3.358406755496332e-06,
+      "loss": 0.7783,
+      "step": 17781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.034072452746175,
+      "learning_rate": 3.3563191707356835e-06,
+      "loss": 1.0684,
+      "step": 17782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.716951087734869,
+      "learning_rate": 3.354232188309861e-06,
+      "loss": 0.696,
+      "step": 17783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8774806461276405,
+      "learning_rate": 3.3521458082769406e-06,
+      "loss": 0.6699,
+      "step": 17784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.67164209254198,
+      "learning_rate": 3.350060030694993e-06,
+      "loss": 0.8145,
+      "step": 17785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4084089427901714,
+      "learning_rate": 3.347974855622063e-06,
+      "loss": 0.7451,
+      "step": 17786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3707093079568513,
+      "learning_rate": 3.345890283116171e-06,
+      "loss": 0.6462,
+      "step": 17787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.275526627172101,
+      "learning_rate": 3.343806313235337e-06,
+      "loss": 1.2266,
+      "step": 17788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4012922265203827,
+      "learning_rate": 3.3417229460375622e-06,
+      "loss": 0.5562,
+      "step": 17789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9114125657052767,
+      "learning_rate": 3.339640181580822e-06,
+      "loss": 0.9336,
+      "step": 17790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.076253411114112,
+      "learning_rate": 3.3375580199230786e-06,
+      "loss": 0.8096,
+      "step": 17791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5675094410735184,
+      "learning_rate": 3.3354764611222693e-06,
+      "loss": 0.6116,
+      "step": 17792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.3502192146671455,
+      "learning_rate": 3.3333955052363364e-06,
+      "loss": 0.8987,
+      "step": 17793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.315930785711292,
+      "learning_rate": 3.331315152323186e-06,
+      "loss": 1.1797,
+      "step": 17794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7508096456694355,
+      "learning_rate": 3.329235402440714e-06,
+      "loss": 1.2207,
+      "step": 17795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7197730054330032,
+      "learning_rate": 3.3271562556467934e-06,
+      "loss": 1.0479,
+      "step": 17796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.933665942761363,
+      "learning_rate": 3.3250777119992916e-06,
+      "loss": 0.7549,
+      "step": 17797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.920492895931779,
+      "learning_rate": 3.3229997715560567e-06,
+      "loss": 1.2427,
+      "step": 17798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5162047861760892,
+      "learning_rate": 3.320922434374915e-06,
+      "loss": 0.8916,
+      "step": 17799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.614252796531656,
+      "learning_rate": 3.318845700513676e-06,
+      "loss": 1.4746,
+      "step": 17800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.086077853277898,
+      "learning_rate": 3.3167695700301286e-06,
+      "loss": 0.9453,
+      "step": 17801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.810257848935981,
+      "learning_rate": 3.314694042982064e-06,
+      "loss": 1.7734,
+      "step": 17802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.735271104666805,
+      "learning_rate": 3.312619119427235e-06,
+      "loss": 0.9609,
+      "step": 17803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4189879896443,
+      "learning_rate": 3.310544799423387e-06,
+      "loss": 1.3242,
+      "step": 17804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5635551094813374,
+      "learning_rate": 3.3084710830282418e-06,
+      "loss": 0.7441,
+      "step": 17805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9807959060800546,
+      "learning_rate": 3.3063979702995158e-06,
+      "loss": 1.0742,
+      "step": 17806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4298014230790392,
+      "learning_rate": 3.3043254612949104e-06,
+      "loss": 1.2949,
+      "step": 17807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8389180811679,
+      "learning_rate": 3.3022535560720908e-06,
+      "loss": 0.999,
+      "step": 17808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1833143243132036,
+      "learning_rate": 3.300182254688719e-06,
+      "loss": 0.5773,
+      "step": 17809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1851862569391516,
+      "learning_rate": 3.298111557202446e-06,
+      "loss": 1.709,
+      "step": 17810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1759117539337347,
+      "learning_rate": 3.296041463670893e-06,
+      "loss": 1.252,
+      "step": 17811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.208940230652798,
+      "learning_rate": 3.2939719741516717e-06,
+      "loss": 0.7808,
+      "step": 17812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4716118750883767,
+      "learning_rate": 3.2919030887023665e-06,
+      "loss": 0.8403,
+      "step": 17813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.673568631874983,
+      "learning_rate": 3.289834807380568e-06,
+      "loss": 1.7676,
+      "step": 17814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3406109202046306,
+      "learning_rate": 3.287767130243824e-06,
+      "loss": 1.2959,
+      "step": 17815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.628939742596057,
+      "learning_rate": 3.285700057349686e-06,
+      "loss": 1.3984,
+      "step": 17816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.737920848947931,
+      "learning_rate": 3.2836335887556746e-06,
+      "loss": 0.9375,
+      "step": 17817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2365821660770733,
+      "learning_rate": 3.2815677245192937e-06,
+      "loss": 1.1719,
+      "step": 17818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.538278926647297,
+      "learning_rate": 3.2795024646980506e-06,
+      "loss": 0.7856,
+      "step": 17819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0684165478406085,
+      "learning_rate": 3.277437809349407e-06,
+      "loss": 1.1543,
+      "step": 17820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.775000419273001,
+      "learning_rate": 3.2753737585308282e-06,
+      "loss": 0.7974,
+      "step": 17821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.965158156861559,
+      "learning_rate": 3.2733103122997486e-06,
+      "loss": 1.4922,
+      "step": 17822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.478072678978412,
+      "learning_rate": 3.271247470713604e-06,
+      "loss": 0.5327,
+      "step": 17823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.274236689358888,
+      "learning_rate": 3.2691852338297883e-06,
+      "loss": 0.6655,
+      "step": 17824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.839310586825519,
+      "learning_rate": 3.267123601705707e-06,
+      "loss": 0.7725,
+      "step": 17825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.163851791690485,
+      "learning_rate": 3.265062574398725e-06,
+      "loss": 0.709,
+      "step": 17826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4659609192385545,
+      "learning_rate": 3.2630021519662073e-06,
+      "loss": 0.9619,
+      "step": 17827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.691015295415452,
+      "learning_rate": 3.2609423344654916e-06,
+      "loss": 1.6211,
+      "step": 17828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6962855793824407,
+      "learning_rate": 3.2588831219539013e-06,
+      "loss": 1.2852,
+      "step": 17829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.11725506551553,
+      "learning_rate": 3.256824514488735e-06,
+      "loss": 1.1709,
+      "step": 17830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.939208163069518,
+      "learning_rate": 3.254766512127297e-06,
+      "loss": 0.6895,
+      "step": 17831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4969609862836672,
+      "learning_rate": 3.2527091149268552e-06,
+      "loss": 1.1221,
+      "step": 17832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7769616316169907,
+      "learning_rate": 3.2506523229446638e-06,
+      "loss": 1.2129,
+      "step": 17833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.806036300798778,
+      "learning_rate": 3.248596136237961e-06,
+      "loss": 0.9487,
+      "step": 17834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.128962692772421,
+      "learning_rate": 3.246540554863972e-06,
+      "loss": 1.5488,
+      "step": 17835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.124645989045203,
+      "learning_rate": 3.2444855788799076e-06,
+      "loss": 1.2051,
+      "step": 17836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.689074033172718,
+      "learning_rate": 3.2424312083429554e-06,
+      "loss": 1.459,
+      "step": 17837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.680442133341455,
+      "learning_rate": 3.240377443310283e-06,
+      "loss": 0.7314,
+      "step": 17838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.178401882354898,
+      "learning_rate": 3.2383242838390447e-06,
+      "loss": 0.9561,
+      "step": 17839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1147798594084835,
+      "learning_rate": 3.2362717299863844e-06,
+      "loss": 0.7754,
+      "step": 17840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9006732652975975,
+      "learning_rate": 3.234219781809425e-06,
+      "loss": 0.9707,
+      "step": 17841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.090500568452802,
+      "learning_rate": 3.2321684393652657e-06,
+      "loss": 0.7832,
+      "step": 17842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.986852286017878,
+      "learning_rate": 3.2301177027109887e-06,
+      "loss": 0.7559,
+      "step": 17843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9365374112069904,
+      "learning_rate": 3.228067571903684e-06,
+      "loss": 0.3893,
+      "step": 17844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3567294327121506,
+      "learning_rate": 3.226018047000395e-06,
+      "loss": 1.1104,
+      "step": 17845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.499119354543805,
+      "learning_rate": 3.2239691280581575e-06,
+      "loss": 0.5928,
+      "step": 17846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.697409854732753,
+      "learning_rate": 3.2219208151339915e-06,
+      "loss": 1.4453,
+      "step": 17847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8257192421907833,
+      "learning_rate": 3.219873108284907e-06,
+      "loss": 1.1631,
+      "step": 17848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3898028946010426,
+      "learning_rate": 3.2178260075678894e-06,
+      "loss": 1.0469,
+      "step": 17849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2354284617411166,
+      "learning_rate": 3.2157795130399045e-06,
+      "loss": 1.2734,
+      "step": 17850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4569275171552056,
+      "learning_rate": 3.213733624757903e-06,
+      "loss": 1.3008,
+      "step": 17851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9644279741302793,
+      "learning_rate": 3.2116883427788235e-06,
+      "loss": 1.0713,
+      "step": 17852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.671275377761195,
+      "learning_rate": 3.2096436671595952e-06,
+      "loss": 1.5293,
+      "step": 17853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5613358807442723,
+      "learning_rate": 3.2075995979571118e-06,
+      "loss": 1.2891,
+      "step": 17854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.620462492242217,
+      "learning_rate": 3.2055561352282602e-06,
+      "loss": 1.0049,
+      "step": 17855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2511296085783066,
+      "learning_rate": 3.2035132790299014e-06,
+      "loss": 0.814,
+      "step": 17856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6565404396590586,
+      "learning_rate": 3.2014710294189035e-06,
+      "loss": 0.7007,
+      "step": 17857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.21874815974376,
+      "learning_rate": 3.1994293864520906e-06,
+      "loss": 1.3066,
+      "step": 17858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.665382399077236,
+      "learning_rate": 3.197388350186284e-06,
+      "loss": 0.8745,
+      "step": 17859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.852519382673119,
+      "learning_rate": 3.1953479206782768e-06,
+      "loss": 0.9053,
+      "step": 17860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6578779729440702,
+      "learning_rate": 3.1933080979848677e-06,
+      "loss": 1.1465,
+      "step": 17861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4969162996049077,
+      "learning_rate": 3.1912688821628093e-06,
+      "loss": 0.9717,
+      "step": 17862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3258751360440693,
+      "learning_rate": 3.1892302732688672e-06,
+      "loss": 0.9551,
+      "step": 17863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.737003567069938,
+      "learning_rate": 3.1871922713597596e-06,
+      "loss": 1.208,
+      "step": 17864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2170420439169116,
+      "learning_rate": 3.185154876492216e-06,
+      "loss": 1.0024,
+      "step": 17865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2161371669457886,
+      "learning_rate": 3.1831180887229333e-06,
+      "loss": 1.3276,
+      "step": 17866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.147682421232956,
+      "learning_rate": 3.1810819081085934e-06,
+      "loss": 0.5874,
+      "step": 17867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.167402481038393,
+      "learning_rate": 3.1790463347058543e-06,
+      "loss": 1.707,
+      "step": 17868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.205837879106569,
+      "learning_rate": 3.1770113685713783e-06,
+      "loss": 0.7776,
+      "step": 17869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.944837994882652,
+      "learning_rate": 3.1749770097617927e-06,
+      "loss": 0.7456,
+      "step": 17870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.877277008914956,
+      "learning_rate": 3.172943258333705e-06,
+      "loss": 0.7427,
+      "step": 17871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9668040270708467,
+      "learning_rate": 3.170910114343728e-06,
+      "loss": 1.0674,
+      "step": 17872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.578704068273491,
+      "learning_rate": 3.168877577848428e-06,
+      "loss": 0.9209,
+      "step": 17873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7827575541088647,
+      "learning_rate": 3.1668456489043846e-06,
+      "loss": 0.9453,
+      "step": 17874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1343106887087173,
+      "learning_rate": 3.1648143275681386e-06,
+      "loss": 1.2285,
+      "step": 17875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6018018663677647,
+      "learning_rate": 3.1627836138962195e-06,
+      "loss": 0.5811,
+      "step": 17876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.132894648581008,
+      "learning_rate": 3.1607535079451383e-06,
+      "loss": 1.0322,
+      "step": 17877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.899724874928786,
+      "learning_rate": 3.1587240097714023e-06,
+      "loss": 1.0859,
+      "step": 17878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3425209424797218,
+      "learning_rate": 3.1566951194314824e-06,
+      "loss": 0.5107,
+      "step": 17879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.666743595760776,
+      "learning_rate": 3.154666836981843e-06,
+      "loss": 1.3262,
+      "step": 17880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.79294800679028,
+      "learning_rate": 3.152639162478935e-06,
+      "loss": 1.0342,
+      "step": 17881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.275479071213781,
+      "learning_rate": 3.150612095979183e-06,
+      "loss": 0.9316,
+      "step": 17882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.505587341465418,
+      "learning_rate": 3.148585637539003e-06,
+      "loss": 0.8008,
+      "step": 17883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4993804655800576,
+      "learning_rate": 3.1465597872147923e-06,
+      "loss": 0.9712,
+      "step": 17884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5409330549612914,
+      "learning_rate": 3.144534545062919e-06,
+      "loss": 0.8477,
+      "step": 17885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.151675361105532,
+      "learning_rate": 3.142509911139757e-06,
+      "loss": 1.1504,
+      "step": 17886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.165658410545341,
+      "learning_rate": 3.1404858855016457e-06,
+      "loss": 0.6318,
+      "step": 17887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.629095243601258,
+      "learning_rate": 3.138462468204914e-06,
+      "loss": 1.0903,
+      "step": 17888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.412522574503208,
+      "learning_rate": 3.1364396593058675e-06,
+      "loss": 1.3682,
+      "step": 17889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.17960841537898,
+      "learning_rate": 3.1344174588608028e-06,
+      "loss": 0.5508,
+      "step": 17890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1682134832449407,
+      "learning_rate": 3.132395866926005e-06,
+      "loss": 0.9692,
+      "step": 17891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0686763840059545,
+      "learning_rate": 3.1303748835577294e-06,
+      "loss": 0.7812,
+      "step": 17892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4211384299218586,
+      "learning_rate": 3.1283545088122145e-06,
+      "loss": 0.7352,
+      "step": 17893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8514927685716316,
+      "learning_rate": 3.1263347427456872e-06,
+      "loss": 0.7656,
+      "step": 17894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6706392846721823,
+      "learning_rate": 3.1243155854143643e-06,
+      "loss": 1.2148,
+      "step": 17895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4790104509258617,
+      "learning_rate": 3.1222970368744302e-06,
+      "loss": 1.3125,
+      "step": 17896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.036035271007449,
+      "learning_rate": 3.1202790971820662e-06,
+      "loss": 0.8096,
+      "step": 17897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3073150656541954,
+      "learning_rate": 3.118261766393421e-06,
+      "loss": 0.8975,
+      "step": 17898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5336231362905623,
+      "learning_rate": 3.1162450445646417e-06,
+      "loss": 1.4258,
+      "step": 17899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9777392366708146,
+      "learning_rate": 3.1142289317518614e-06,
+      "loss": 1.002,
+      "step": 17900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.303092842342503,
+      "learning_rate": 3.112213428011179e-06,
+      "loss": 0.6611,
+      "step": 17901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.729885579008439,
+      "learning_rate": 3.1101985333986865e-06,
+      "loss": 0.79,
+      "step": 17902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6645917171041162,
+      "learning_rate": 3.1081842479704527e-06,
+      "loss": 1.0332,
+      "step": 17903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7934981034690063,
+      "learning_rate": 3.1061705717825463e-06,
+      "loss": 1.7891,
+      "step": 17904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.751935664849243,
+      "learning_rate": 3.1041575048909976e-06,
+      "loss": 1.5254,
+      "step": 17905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.870323851537252,
+      "learning_rate": 3.102145047351826e-06,
+      "loss": 1.7363,
+      "step": 17906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4445714833260785,
+      "learning_rate": 3.1001331992210477e-06,
+      "loss": 1.4648,
+      "step": 17907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5189167803035155,
+      "learning_rate": 3.0981219605546453e-06,
+      "loss": 0.8936,
+      "step": 17908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.847935597058057,
+      "learning_rate": 3.096111331408594e-06,
+      "loss": 0.8271,
+      "step": 17909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.141008217472875,
+      "learning_rate": 3.0941013118388484e-06,
+      "loss": 0.9033,
+      "step": 17910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2147645442285664,
+      "learning_rate": 3.0920919019013423e-06,
+      "loss": 1.2227,
+      "step": 17911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.03802836694716,
+      "learning_rate": 3.0900831016520026e-06,
+      "loss": 1.0703,
+      "step": 17912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.286657634147447,
+      "learning_rate": 3.0880749111467322e-06,
+      "loss": 2.252,
+      "step": 17913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.657194351174895,
+      "learning_rate": 3.086067330441414e-06,
+      "loss": 0.5199,
+      "step": 17914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4743298596270353,
+      "learning_rate": 3.08406035959192e-06,
+      "loss": 1.0723,
+      "step": 17915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.617032585837102,
+      "learning_rate": 3.0820539986541054e-06,
+      "loss": 0.9629,
+      "step": 17916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.989520684803121,
+      "learning_rate": 3.0800482476838003e-06,
+      "loss": 1.1895,
+      "step": 17917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4421003896025093,
+      "learning_rate": 3.078043106736836e-06,
+      "loss": 0.7891,
+      "step": 17918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.5974405816955946,
+      "learning_rate": 3.0760385758690085e-06,
+      "loss": 1.1221,
+      "step": 17919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6754413917950393,
+      "learning_rate": 3.0740346551360934e-06,
+      "loss": 0.3926,
+      "step": 17920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0514864626388305,
+      "learning_rate": 3.0720313445938757e-06,
+      "loss": 0.6929,
+      "step": 17921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6551276135711714,
+      "learning_rate": 3.070028644298098e-06,
+      "loss": 1.0098,
+      "step": 17922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9202827122289192,
+      "learning_rate": 3.0680265543044957e-06,
+      "loss": 1.0576,
+      "step": 17923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5002069330528973,
+      "learning_rate": 3.0660250746687796e-06,
+      "loss": 0.8335,
+      "step": 17924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.578594092595028,
+      "learning_rate": 3.064024205446664e-06,
+      "loss": 1.8008,
+      "step": 17925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6321231833911507,
+      "learning_rate": 3.0620239466938177e-06,
+      "loss": 1.5098,
+      "step": 17926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5534196206423765,
+      "learning_rate": 3.0600242984659187e-06,
+      "loss": 1.1934,
+      "step": 17927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9825226909903386,
+      "learning_rate": 3.058025260818609e-06,
+      "loss": 1.2119,
+      "step": 17928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.851866298813451,
+      "learning_rate": 3.0560268338075264e-06,
+      "loss": 1.0771,
+      "step": 17929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.940868213004261,
+      "learning_rate": 3.054029017488283e-06,
+      "loss": 0.7637,
+      "step": 17930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8482049687731879,
+      "learning_rate": 3.052031811916481e-06,
+      "loss": 0.8362,
+      "step": 17931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.8095728569134195,
+      "learning_rate": 3.0500352171476897e-06,
+      "loss": 1.4629,
+      "step": 17932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.176493064282174,
+      "learning_rate": 3.04803923323749e-06,
+      "loss": 0.729,
+      "step": 17933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.612168529147625,
+      "learning_rate": 3.046043860241421e-06,
+      "loss": 0.7383,
+      "step": 17934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1552457820637336,
+      "learning_rate": 3.0440490982150155e-06,
+      "loss": 0.9277,
+      "step": 17935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6508552071001557,
+      "learning_rate": 3.0420549472137795e-06,
+      "loss": 1.1807,
+      "step": 17936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.173216391513834,
+      "learning_rate": 3.0400614072932125e-06,
+      "loss": 1.2129,
+      "step": 17937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.231426718781269,
+      "learning_rate": 3.038068478508804e-06,
+      "loss": 0.8672,
+      "step": 17938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.919915431166074,
+      "learning_rate": 3.0360761609160098e-06,
+      "loss": 1.54,
+      "step": 17939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3569304332371486,
+      "learning_rate": 3.034084454570274e-06,
+      "loss": 0.9053,
+      "step": 17940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.266047984203482,
+      "learning_rate": 3.032093359527019e-06,
+      "loss": 1.3125,
+      "step": 17941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1932574118525157,
+      "learning_rate": 3.03010287584167e-06,
+      "loss": 1.2988,
+      "step": 17942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9480031318699673,
+      "learning_rate": 3.0281130035696133e-06,
+      "loss": 1.5273,
+      "step": 17943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.574928475284708,
+      "learning_rate": 3.026123742766221e-06,
+      "loss": 1.0938,
+      "step": 17944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.97246899908992,
+      "learning_rate": 3.0241350934868603e-06,
+      "loss": 1.0254,
+      "step": 17945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.025992294010672,
+      "learning_rate": 3.022147055786878e-06,
+      "loss": 1.3223,
+      "step": 17946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6708489237348914,
+      "learning_rate": 3.0201596297215945e-06,
+      "loss": 0.7114,
+      "step": 17947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.027338693131037,
+      "learning_rate": 3.0181728153463233e-06,
+      "loss": 1.4727,
+      "step": 17948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6154777749761213,
+      "learning_rate": 3.0161866127163486e-06,
+      "loss": 0.8945,
+      "step": 17949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8484413100911334,
+      "learning_rate": 3.0142010218869533e-06,
+      "loss": 1.4902,
+      "step": 17950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.456271847867981,
+      "learning_rate": 3.012216042913396e-06,
+      "loss": 0.1864,
+      "step": 17951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.258967754125425,
+      "learning_rate": 3.0102316758509134e-06,
+      "loss": 1.0566,
+      "step": 17952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7994978628185625,
+      "learning_rate": 3.0082479207547254e-06,
+      "loss": 0.4612,
+      "step": 17953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8593661524457963,
+      "learning_rate": 3.006264777680046e-06,
+      "loss": 1.6074,
+      "step": 17954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7171893010369614,
+      "learning_rate": 3.004282246682069e-06,
+      "loss": 1.0176,
+      "step": 17955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.577242717418975,
+      "learning_rate": 3.002300327815963e-06,
+      "loss": 1.4219,
+      "step": 17956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4914796595316333,
+      "learning_rate": 3.0003190211368854e-06,
+      "loss": 0.979,
+      "step": 17957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.527142186283153,
+      "learning_rate": 2.998338326699965e-06,
+      "loss": 0.5547,
+      "step": 17958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0821881261988073,
+      "learning_rate": 2.9963582445603406e-06,
+      "loss": 1.0928,
+      "step": 17959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.137485586186466,
+      "learning_rate": 2.9943787747731043e-06,
+      "loss": 0.8643,
+      "step": 17960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9258310802891667,
+      "learning_rate": 2.9923999173933515e-06,
+      "loss": 0.5181,
+      "step": 17961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1183837664123364,
+      "learning_rate": 2.990421672476143e-06,
+      "loss": 0.9501,
+      "step": 17962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.4664541896177665,
+      "learning_rate": 2.9884440400765406e-06,
+      "loss": 0.9395,
+      "step": 17963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.756426930533505,
+      "learning_rate": 2.9864670202495837e-06,
+      "loss": 1.0596,
+      "step": 17964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.872134613695493,
+      "learning_rate": 2.9844906130502893e-06,
+      "loss": 0.8965,
+      "step": 17965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7470099059080106,
+      "learning_rate": 2.982514818533652e-06,
+      "loss": 1.1406,
+      "step": 17966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.884109660102081,
+      "learning_rate": 2.9805396367546673e-06,
+      "loss": 1.1357,
+      "step": 17967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4045438786796653,
+      "learning_rate": 2.978565067768305e-06,
+      "loss": 0.7773,
+      "step": 17968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.726906814503376,
+      "learning_rate": 2.976591111629509e-06,
+      "loss": 1.2129,
+      "step": 17969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9997544213249148,
+      "learning_rate": 2.974617768393212e-06,
+      "loss": 0.9307,
+      "step": 17970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8520952576924585,
+      "learning_rate": 2.9726450381143412e-06,
+      "loss": 1.3398,
+      "step": 17971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.395453679230922,
+      "learning_rate": 2.9706729208477917e-06,
+      "loss": 0.9404,
+      "step": 17972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4296603886047596,
+      "learning_rate": 2.9687014166484394e-06,
+      "loss": 1.0176,
+      "step": 17973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.248416661876707,
+      "learning_rate": 2.9667305255711653e-06,
+      "loss": 0.8779,
+      "step": 17974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.526298900482895,
+      "learning_rate": 2.964760247670806e-06,
+      "loss": 1.4746,
+      "step": 17975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1477724040157207,
+      "learning_rate": 2.9627905830022036e-06,
+      "loss": 1.3428,
+      "step": 17976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.489863923885511,
+      "learning_rate": 2.960821531620167e-06,
+      "loss": 0.5635,
+      "step": 17977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8563810422790943,
+      "learning_rate": 2.9588530935794944e-06,
+      "loss": 1.2793,
+      "step": 17978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4037413958684635,
+      "learning_rate": 2.9568852689349642e-06,
+      "loss": 1.3457,
+      "step": 17979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.421175532507936,
+      "learning_rate": 2.954918057741346e-06,
+      "loss": 0.9092,
+      "step": 17980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9714779818272885,
+      "learning_rate": 2.952951460053385e-06,
+      "loss": 0.7852,
+      "step": 17981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3555101030175183,
+      "learning_rate": 2.950985475925805e-06,
+      "loss": 0.4214,
+      "step": 17982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1262287626377634,
+      "learning_rate": 2.949020105413322e-06,
+      "loss": 0.5818,
+      "step": 17983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.011509678911462,
+      "learning_rate": 2.9470553485706376e-06,
+      "loss": 0.6475,
+      "step": 17984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.251103978057352,
+      "learning_rate": 2.945091205452424e-06,
+      "loss": 0.8301,
+      "step": 17985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3743739055762565,
+      "learning_rate": 2.943127676113347e-06,
+      "loss": 0.585,
+      "step": 17986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.254155983138607,
+      "learning_rate": 2.9411647606080396e-06,
+      "loss": 1.2734,
+      "step": 17987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.063503801530559,
+      "learning_rate": 2.9392024589911416e-06,
+      "loss": 1.1836,
+      "step": 17988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0395399681938944,
+      "learning_rate": 2.9372407713172594e-06,
+      "loss": 0.6497,
+      "step": 17989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.367842077015393,
+      "learning_rate": 2.935279697640986e-06,
+      "loss": 1.1572,
+      "step": 17990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.650770482031023,
+      "learning_rate": 2.9333192380168878e-06,
+      "loss": 1.1377,
+      "step": 17991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.840666611944974,
+      "learning_rate": 2.9313593924995356e-06,
+      "loss": 0.8657,
+      "step": 17992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.876709714690423,
+      "learning_rate": 2.9294001611434695e-06,
+      "loss": 1.793,
+      "step": 17993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7835359073253185,
+      "learning_rate": 2.927441544003212e-06,
+      "loss": 1.1191,
+      "step": 17994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9791335320627017,
+      "learning_rate": 2.9254835411332727e-06,
+      "loss": 0.8428,
+      "step": 17995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.289705153728209,
+      "learning_rate": 2.9235261525881323e-06,
+      "loss": 0.9021,
+      "step": 17996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3750057220390053,
+      "learning_rate": 2.9215693784222787e-06,
+      "loss": 0.8418,
+      "step": 17997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3166122821841073,
+      "learning_rate": 2.919613218690162e-06,
+      "loss": 1.3672,
+      "step": 17998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2341504817814175,
+      "learning_rate": 2.9176576734462196e-06,
+      "loss": 1.085,
+      "step": 17999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.317446033112463,
+      "learning_rate": 2.9157027427448663e-06,
+      "loss": 0.876,
+      "step": 18000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5128477941980925,
+      "learning_rate": 2.91374842664052e-06,
+      "loss": 0.4908,
+      "step": 18001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0671359414942723,
+      "learning_rate": 2.9117947251875666e-06,
+      "loss": 1.3555,
+      "step": 18002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2675267097893586,
+      "learning_rate": 2.909841638440372e-06,
+      "loss": 0.9211,
+      "step": 18003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.257722906167465,
+      "learning_rate": 2.9078891664532897e-06,
+      "loss": 1.1396,
+      "step": 18004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2794375600929606,
+      "learning_rate": 2.9059373092806624e-06,
+      "loss": 1.2676,
+      "step": 18005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1280531654912993,
+      "learning_rate": 2.903986066976805e-06,
+      "loss": 0.8721,
+      "step": 18006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.468020947161892,
+      "learning_rate": 2.902035439596018e-06,
+      "loss": 1.0312,
+      "step": 18007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.277745997461483,
+      "learning_rate": 2.9000854271925867e-06,
+      "loss": 1.3984,
+      "step": 18008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9149525282905553,
+      "learning_rate": 2.8981360298207834e-06,
+      "loss": 0.7437,
+      "step": 18009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2382047683404602,
+      "learning_rate": 2.8961872475348538e-06,
+      "loss": 1.2227,
+      "step": 18010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8839227718917435,
+      "learning_rate": 2.8942390803890385e-06,
+      "loss": 1.4141,
+      "step": 18011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.043568977015055,
+      "learning_rate": 2.892291528437552e-06,
+      "loss": 1.2153,
+      "step": 18012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.245131228230461,
+      "learning_rate": 2.890344591734584e-06,
+      "loss": 0.7275,
+      "step": 18013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.996635229195447,
+      "learning_rate": 2.8883982703343332e-06,
+      "loss": 1.2637,
+      "step": 18014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.322404528442814,
+      "learning_rate": 2.8864525642909563e-06,
+      "loss": 1.2109,
+      "step": 18015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.483120940635342,
+      "learning_rate": 2.8845074736586016e-06,
+      "loss": 1.293,
+      "step": 18016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2472469700534767,
+      "learning_rate": 2.8825629984913947e-06,
+      "loss": 0.9668,
+      "step": 18017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.886478385067441,
+      "learning_rate": 2.880619138843463e-06,
+      "loss": 1.2578,
+      "step": 18018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2013132207349484,
+      "learning_rate": 2.8786758947688896e-06,
+      "loss": 1.3618,
+      "step": 18019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.074016209140808,
+      "learning_rate": 2.8767332663217655e-06,
+      "loss": 1.5332,
+      "step": 18020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.924196135638626,
+      "learning_rate": 2.874791253556142e-06,
+      "loss": 0.9229,
+      "step": 18021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.615271908841947,
+      "learning_rate": 2.8728498565260776e-06,
+      "loss": 0.9492,
+      "step": 18022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.672251803057506,
+      "learning_rate": 2.870909075285594e-06,
+      "loss": 0.6211,
+      "step": 18023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8484622415158867,
+      "learning_rate": 2.8689689098887004e-06,
+      "loss": 0.9941,
+      "step": 18024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.082776027488753,
+      "learning_rate": 2.8670293603893866e-06,
+      "loss": 1.4922,
+      "step": 18025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.270970610606568,
+      "learning_rate": 2.8650904268416434e-06,
+      "loss": 0.8613,
+      "step": 18026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5032379163021625,
+      "learning_rate": 2.8631521092994217e-06,
+      "loss": 0.8369,
+      "step": 18027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.641928537267163,
+      "learning_rate": 2.8612144078166593e-06,
+      "loss": 0.3761,
+      "step": 18028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9390895682046163,
+      "learning_rate": 2.85927732244729e-06,
+      "loss": 1.4629,
+      "step": 18029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7345836373628063,
+      "learning_rate": 2.857340853245216e-06,
+      "loss": 0.9038,
+      "step": 18030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.625897088554972,
+      "learning_rate": 2.8554050002643384e-06,
+      "loss": 1.1143,
+      "step": 18031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4541856938797326,
+      "learning_rate": 2.853469763558522e-06,
+      "loss": 0.5938,
+      "step": 18032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.562043049543556,
+      "learning_rate": 2.85153514318163e-06,
+      "loss": 1.3516,
+      "step": 18033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0801849443469846,
+      "learning_rate": 2.849601139187491e-06,
+      "loss": 0.8855,
+      "step": 18034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2162459306415996,
+      "learning_rate": 2.84766775162994e-06,
+      "loss": 1.7422,
+      "step": 18035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.97953987896115,
+      "learning_rate": 2.8457349805627754e-06,
+      "loss": 1.0771,
+      "step": 18036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4376617393590267,
+      "learning_rate": 2.8438028260397903e-06,
+      "loss": 1.1729,
+      "step": 18037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8853496037781596,
+      "learning_rate": 2.841871288114745e-06,
+      "loss": 1.8652,
+      "step": 18038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3126495931542577,
+      "learning_rate": 2.8399403668414038e-06,
+      "loss": 0.8701,
+      "step": 18039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.505001679313254,
+      "learning_rate": 2.8380100622735023e-06,
+      "loss": 1.2852,
+      "step": 18040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.364447591193136,
+      "learning_rate": 2.8360803744647615e-06,
+      "loss": 0.9307,
+      "step": 18041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.103844678630006,
+      "learning_rate": 2.8341513034688767e-06,
+      "loss": 0.8699,
+      "step": 18042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.678842483611523,
+      "learning_rate": 2.8322228493395416e-06,
+      "loss": 0.4274,
+      "step": 18043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0233901855551704,
+      "learning_rate": 2.8302950121304184e-06,
+      "loss": 1.085,
+      "step": 18044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5054766748515336,
+      "learning_rate": 2.828367791895162e-06,
+      "loss": 0.8682,
+      "step": 18045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.018497283517948,
+      "learning_rate": 2.8264411886873953e-06,
+      "loss": 1.125,
+      "step": 18046
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.860292381333396,
+      "learning_rate": 2.8245152025607457e-06,
+      "loss": 1.0078,
+      "step": 18047
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.437343025134025,
+      "learning_rate": 2.822589833568817e-06,
+      "loss": 1.1777,
+      "step": 18048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2606232956480956,
+      "learning_rate": 2.8206650817651835e-06,
+      "loss": 0.9395,
+      "step": 18049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2114365832573575,
+      "learning_rate": 2.8187409472034105e-06,
+      "loss": 1.0703,
+      "step": 18050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8812058427767384,
+      "learning_rate": 2.8168174299370436e-06,
+      "loss": 0.5415,
+      "step": 18051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.75920834756298,
+      "learning_rate": 2.8148945300196213e-06,
+      "loss": 1.3574,
+      "step": 18052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1409285949602013,
+      "learning_rate": 2.8129722475046527e-06,
+      "loss": 0.8994,
+      "step": 18053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.802968476960514,
+      "learning_rate": 2.8110505824456344e-06,
+      "loss": 0.8457,
+      "step": 18054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9089206685707198,
+      "learning_rate": 2.8091295348960377e-06,
+      "loss": 1.3301,
+      "step": 18055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9925304049084778,
+      "learning_rate": 2.8072091049093364e-06,
+      "loss": 1.1123,
+      "step": 18056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8060445695633844,
+      "learning_rate": 2.8052892925389736e-06,
+      "loss": 0.9434,
+      "step": 18057
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.881997588248089,
+      "learning_rate": 2.8033700978383733e-06,
+      "loss": 1.2637,
+      "step": 18058
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0390382962599065,
+      "learning_rate": 2.801451520860948e-06,
+      "loss": 1.0957,
+      "step": 18059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0635053580392757,
+      "learning_rate": 2.7995335616600866e-06,
+      "loss": 0.9124,
+      "step": 18060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.885047108548226,
+      "learning_rate": 2.7976162202891704e-06,
+      "loss": 1.6035,
+      "step": 18061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.400503929042804,
+      "learning_rate": 2.795699496801557e-06,
+      "loss": 1.3887,
+      "step": 18062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.735963619177569,
+      "learning_rate": 2.7937833912505844e-06,
+      "loss": 0.9619,
+      "step": 18063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.128064293515278,
+      "learning_rate": 2.7918679036895823e-06,
+      "loss": 1.0518,
+      "step": 18064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.682721767112943,
+      "learning_rate": 2.789953034171852e-06,
+      "loss": 1.0369,
+      "step": 18065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.980294358737351,
+      "learning_rate": 2.7880387827506903e-06,
+      "loss": 1.2158,
+      "step": 18066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5769681242280345,
+      "learning_rate": 2.7861251494793685e-06,
+      "loss": 0.6821,
+      "step": 18067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.802379717613804,
+      "learning_rate": 2.784212134411135e-06,
+      "loss": 0.7721,
+      "step": 18068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.790267205566778,
+      "learning_rate": 2.782299737599239e-06,
+      "loss": 1.0745,
+      "step": 18069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.508458704893847,
+      "learning_rate": 2.7803879590968973e-06,
+      "loss": 1.1406,
+      "step": 18070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.397107479354675,
+      "learning_rate": 2.7784767989573108e-06,
+      "loss": 1.4355,
+      "step": 18071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.209351193049906,
+      "learning_rate": 2.7765662572336647e-06,
+      "loss": 1.2715,
+      "step": 18072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1829433393757167,
+      "learning_rate": 2.7746563339791394e-06,
+      "loss": 1.4902,
+      "step": 18073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.146804799538271,
+      "learning_rate": 2.772747029246878e-06,
+      "loss": 1.252,
+      "step": 18074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6753844475114383,
+      "learning_rate": 2.7708383430900126e-06,
+      "loss": 1.0078,
+      "step": 18075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2617881118600964,
+      "learning_rate": 2.7689302755616736e-06,
+      "loss": 0.751,
+      "step": 18076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.8718182223595194,
+      "learning_rate": 2.767022826714949e-06,
+      "loss": 0.9639,
+      "step": 18077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.549313025961031,
+      "learning_rate": 2.765115996602932e-06,
+      "loss": 0.8291,
+      "step": 18078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.835390148463937,
+      "learning_rate": 2.7632097852786864e-06,
+      "loss": 1.3477,
+      "step": 18079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.815634470097444,
+      "learning_rate": 2.7613041927952556e-06,
+      "loss": 0.9873,
+      "step": 18080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.3505568105283485,
+      "learning_rate": 2.7593992192056774e-06,
+      "loss": 2.2656,
+      "step": 18081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.091170343282059,
+      "learning_rate": 2.7574948645629676e-06,
+      "loss": 0.7793,
+      "step": 18082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9403367650233614,
+      "learning_rate": 2.755591128920121e-06,
+      "loss": 0.957,
+      "step": 18083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0184197792159413,
+      "learning_rate": 2.7536880123301105e-06,
+      "loss": 1.4668,
+      "step": 18084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4394029867182887,
+      "learning_rate": 2.7517855148459088e-06,
+      "loss": 1.0264,
+      "step": 18085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.791260009309691,
+      "learning_rate": 2.7498836365204616e-06,
+      "loss": 1.2344,
+      "step": 18086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.561709111103618,
+      "learning_rate": 2.7479823774066966e-06,
+      "loss": 0.9375,
+      "step": 18087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4001505446901685,
+      "learning_rate": 2.746081737557521e-06,
+      "loss": 0.9287,
+      "step": 18088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6145970520227015,
+      "learning_rate": 2.744181717025829e-06,
+      "loss": 0.9639,
+      "step": 18089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5020094425197046,
+      "learning_rate": 2.742282315864503e-06,
+      "loss": 1.002,
+      "step": 18090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.511618344729635,
+      "learning_rate": 2.740383534126398e-06,
+      "loss": 1.1719,
+      "step": 18091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0819578462010595,
+      "learning_rate": 2.738485371864358e-06,
+      "loss": 0.9385,
+      "step": 18092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.879545019291666,
+      "learning_rate": 2.7365878291312016e-06,
+      "loss": 0.8506,
+      "step": 18093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9121780149396346,
+      "learning_rate": 2.734690905979745e-06,
+      "loss": 0.8149,
+      "step": 18094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.613219025763148,
+      "learning_rate": 2.7327946024627794e-06,
+      "loss": 0.7627,
+      "step": 18095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4037634151598177,
+      "learning_rate": 2.7308989186330763e-06,
+      "loss": 0.6709,
+      "step": 18096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.801434343665126,
+      "learning_rate": 2.7290038545433884e-06,
+      "loss": 1.0308,
+      "step": 18097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2416063146963765,
+      "learning_rate": 2.7271094102464533e-06,
+      "loss": 1.3965,
+      "step": 18098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.762110685821998,
+      "learning_rate": 2.7252155857950017e-06,
+      "loss": 1.6777,
+      "step": 18099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3847432254109515,
+      "learning_rate": 2.72332238124173e-06,
+      "loss": 1.498,
+      "step": 18100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.059613889928786,
+      "learning_rate": 2.7214297966393244e-06,
+      "loss": 1.4961,
+      "step": 18101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.301135620391075,
+      "learning_rate": 2.7195378320404555e-06,
+      "loss": 0.9155,
+      "step": 18102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8360386597379943,
+      "learning_rate": 2.717646487497788e-06,
+      "loss": 0.698,
+      "step": 18103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9787706409784516,
+      "learning_rate": 2.715755763063943e-06,
+      "loss": 1.0508,
+      "step": 18104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.761690355552233,
+      "learning_rate": 2.713865658791545e-06,
+      "loss": 0.4756,
+      "step": 18105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9030012369237057,
+      "learning_rate": 2.7119761747331885e-06,
+      "loss": 0.6621,
+      "step": 18106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.010419394591964,
+      "learning_rate": 2.710087310941467e-06,
+      "loss": 1.084,
+      "step": 18107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5386705328865844,
+      "learning_rate": 2.708199067468939e-06,
+      "loss": 1.1553,
+      "step": 18108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.040837476338545,
+      "learning_rate": 2.7063114443681593e-06,
+      "loss": 1.9219,
+      "step": 18109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.438767617072563,
+      "learning_rate": 2.7044244416916498e-06,
+      "loss": 1.3242,
+      "step": 18110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9929435342578967,
+      "learning_rate": 2.702538059491935e-06,
+      "loss": 0.8379,
+      "step": 18111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.236640795043099,
+      "learning_rate": 2.700652297821507e-06,
+      "loss": 1.2559,
+      "step": 18112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.366913892777324,
+      "learning_rate": 2.698767156732848e-06,
+      "loss": 1.168,
+      "step": 18113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4173847157275232,
+      "learning_rate": 2.696882636278425e-06,
+      "loss": 1.1058,
+      "step": 18114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.4153690055308745,
+      "learning_rate": 2.6949987365106703e-06,
+      "loss": 0.9209,
+      "step": 18115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.116879036029204,
+      "learning_rate": 2.6931154574820263e-06,
+      "loss": 0.874,
+      "step": 18116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.815714649081636,
+      "learning_rate": 2.691232799244897e-06,
+      "loss": 0.5769,
+      "step": 18117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.378428544310942,
+      "learning_rate": 2.6893507618516807e-06,
+      "loss": 0.9619,
+      "step": 18118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.192206850190793,
+      "learning_rate": 2.68746934535474e-06,
+      "loss": 1.0222,
+      "step": 18119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.016155202757274,
+      "learning_rate": 2.6855885498064533e-06,
+      "loss": 1.1064,
+      "step": 18120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.866613363383423,
+      "learning_rate": 2.6837083752591445e-06,
+      "loss": 1.0225,
+      "step": 18121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.557432706940627,
+      "learning_rate": 2.6818288217651534e-06,
+      "loss": 1.2559,
+      "step": 18122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.030937053341691,
+      "learning_rate": 2.679949889376776e-06,
+      "loss": 0.7964,
+      "step": 18123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8305186895586676,
+      "learning_rate": 2.678071578146313e-06,
+      "loss": 0.9951,
+      "step": 18124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.4395908211845265,
+      "learning_rate": 2.6761938881260272e-06,
+      "loss": 1.0088,
+      "step": 18125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.819567679420935,
+      "learning_rate": 2.6743168193681807e-06,
+      "loss": 1.8164,
+      "step": 18126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.250025113283515,
+      "learning_rate": 2.6724403719250003e-06,
+      "loss": 0.8066,
+      "step": 18127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.39756674601036,
+      "learning_rate": 2.6705645458487223e-06,
+      "loss": 1.1602,
+      "step": 18128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9120541931286685,
+      "learning_rate": 2.668689341191541e-06,
+      "loss": 1.1016,
+      "step": 18129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1540837117596117,
+      "learning_rate": 2.6668147580056395e-06,
+      "loss": 1.1279,
+      "step": 18130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.0346679450723135,
+      "learning_rate": 2.664940796343196e-06,
+      "loss": 1.1777,
+      "step": 18131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.039179390274943,
+      "learning_rate": 2.6630674562563546e-06,
+      "loss": 1.0137,
+      "step": 18132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.854234787050392,
+      "learning_rate": 2.661194737797257e-06,
+      "loss": 0.5618,
+      "step": 18133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.250312120532826,
+      "learning_rate": 2.659322641018014e-06,
+      "loss": 0.8945,
+      "step": 18134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.457153899462655,
+      "learning_rate": 2.6574511659707287e-06,
+      "loss": 1.832,
+      "step": 18135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2099446470138093,
+      "learning_rate": 2.6555803127074756e-06,
+      "loss": 0.7783,
+      "step": 18136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.676617523714514,
+      "learning_rate": 2.6537100812803308e-06,
+      "loss": 0.7134,
+      "step": 18137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.594232488931799,
+      "learning_rate": 2.6518404717413373e-06,
+      "loss": 1.5547,
+      "step": 18138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.749702164472298,
+      "learning_rate": 2.649971484142527e-06,
+      "loss": 1.7676,
+      "step": 18139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2166584526234883,
+      "learning_rate": 2.6481031185358997e-06,
+      "loss": 0.6045,
+      "step": 18140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.17859526583207,
+      "learning_rate": 2.646235374973474e-06,
+      "loss": 0.7202,
+      "step": 18141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9608856137580997,
+      "learning_rate": 2.6443682535072177e-06,
+      "loss": 2.1621,
+      "step": 18142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7847019548874785,
+      "learning_rate": 2.6425017541890915e-06,
+      "loss": 0.8467,
+      "step": 18143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6802444101735574,
+      "learning_rate": 2.6406358770710375e-06,
+      "loss": 0.9473,
+      "step": 18144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0633443780129945,
+      "learning_rate": 2.638770622204986e-06,
+      "loss": 1.4082,
+      "step": 18145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.185385308587702,
+      "learning_rate": 2.636905989642846e-06,
+      "loss": 0.7192,
+      "step": 18146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8973383297386337,
+      "learning_rate": 2.635041979436509e-06,
+      "loss": 1.5859,
+      "step": 18147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.260111657819265,
+      "learning_rate": 2.633178591637844e-06,
+      "loss": 0.9727,
+      "step": 18148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.508257525448804,
+      "learning_rate": 2.631315826298711e-06,
+      "loss": 1.2412,
+      "step": 18149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1953493596494895,
+      "learning_rate": 2.6294536834709586e-06,
+      "loss": 1.1621,
+      "step": 18150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.967181733701227,
+      "learning_rate": 2.627592163206405e-06,
+      "loss": 1.042,
+      "step": 18151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6593767148216654,
+      "learning_rate": 2.625731265556852e-06,
+      "loss": 0.5215,
+      "step": 18152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8517248656151035,
+      "learning_rate": 2.623870990574084e-06,
+      "loss": 1.459,
+      "step": 18153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8412855455912975,
+      "learning_rate": 2.6220113383098822e-06,
+      "loss": 1.3086,
+      "step": 18154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.698708320223766,
+      "learning_rate": 2.6201523088159957e-06,
+      "loss": 0.874,
+      "step": 18155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.391604059823895,
+      "learning_rate": 2.6182939021441588e-06,
+      "loss": 1.1113,
+      "step": 18156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.315627310795537,
+      "learning_rate": 2.616436118346086e-06,
+      "loss": 1.1436,
+      "step": 18157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9241981464796694,
+      "learning_rate": 2.6145789574734848e-06,
+      "loss": 1.1641,
+      "step": 18158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2516111327616652,
+      "learning_rate": 2.61272241957804e-06,
+      "loss": 1.0723,
+      "step": 18159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.908970300757344,
+      "learning_rate": 2.6108665047114178e-06,
+      "loss": 1.0795,
+      "step": 18160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.45842748127631,
+      "learning_rate": 2.6090112129252636e-06,
+      "loss": 0.8818,
+      "step": 18161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3631266898063665,
+      "learning_rate": 2.6071565442712137e-06,
+      "loss": 1.0762,
+      "step": 18162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6122652596323617,
+      "learning_rate": 2.605302498800882e-06,
+      "loss": 1.043,
+      "step": 18163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.440363546866582,
+      "learning_rate": 2.603449076565867e-06,
+      "loss": 0.7095,
+      "step": 18164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8813492697869783,
+      "learning_rate": 2.6015962776177404e-06,
+      "loss": 1.3418,
+      "step": 18165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2423214804301095,
+      "learning_rate": 2.5997441020080756e-06,
+      "loss": 0.4374,
+      "step": 18166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.976920649457953,
+      "learning_rate": 2.5978925497884082e-06,
+      "loss": 0.748,
+      "step": 18167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.112765809718664,
+      "learning_rate": 2.5960416210102757e-06,
+      "loss": 0.6914,
+      "step": 18168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9511773736549773,
+      "learning_rate": 2.5941913157251835e-06,
+      "loss": 0.6904,
+      "step": 18169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.769892616531677,
+      "learning_rate": 2.592341633984624e-06,
+      "loss": 0.7373,
+      "step": 18170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.680144652135909,
+      "learning_rate": 2.5904925758400757e-06,
+      "loss": 0.7461,
+      "step": 18171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2857823631389587,
+      "learning_rate": 2.5886441413429997e-06,
+      "loss": 0.6816,
+      "step": 18172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8629854555284042,
+      "learning_rate": 2.58679633054483e-06,
+      "loss": 1.0743,
+      "step": 18173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8395040729018275,
+      "learning_rate": 2.5849491434969924e-06,
+      "loss": 0.916,
+      "step": 18174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2219909351110068,
+      "learning_rate": 2.5831025802508983e-06,
+      "loss": 0.6543,
+      "step": 18175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1529986428886847,
+      "learning_rate": 2.581256640857932e-06,
+      "loss": 1.5547,
+      "step": 18176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.785273677375946,
+      "learning_rate": 2.5794113253694657e-06,
+      "loss": 1.6543,
+      "step": 18177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.339288152029916,
+      "learning_rate": 2.5775666338368504e-06,
+      "loss": 0.8779,
+      "step": 18178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1468936436272568,
+      "learning_rate": 2.5757225663114363e-06,
+      "loss": 0.6772,
+      "step": 18179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.804142976144041,
+      "learning_rate": 2.5738791228445303e-06,
+      "loss": 1.6602,
+      "step": 18180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.109905623667435,
+      "learning_rate": 2.5720363034874405e-06,
+      "loss": 1.1504,
+      "step": 18181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.503737897762735,
+      "learning_rate": 2.570194108291446e-06,
+      "loss": 1.2939,
+      "step": 18182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1096786845558,
+      "learning_rate": 2.5683525373078223e-06,
+      "loss": 0.8535,
+      "step": 18183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3042138598979904,
+      "learning_rate": 2.566511590587817e-06,
+      "loss": 0.8311,
+      "step": 18184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.985576777889802,
+      "learning_rate": 2.5646712681826586e-06,
+      "loss": 0.9854,
+      "step": 18185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.5510109249982165,
+      "learning_rate": 2.5628315701435617e-06,
+      "loss": 1.1514,
+      "step": 18186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.603986647423703,
+      "learning_rate": 2.5609924965217304e-06,
+      "loss": 1.0986,
+      "step": 18187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.213602952861168,
+      "learning_rate": 2.5591540473683453e-06,
+      "loss": 1.0947,
+      "step": 18188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.071058442924042,
+      "learning_rate": 2.557316222734568e-06,
+      "loss": 0.6045,
+      "step": 18189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.3517962589861385,
+      "learning_rate": 2.5554790226715445e-06,
+      "loss": 0.9395,
+      "step": 18190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9330835150859358,
+      "learning_rate": 2.553642447230395e-06,
+      "loss": 0.2112,
+      "step": 18191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4574480779679155,
+      "learning_rate": 2.551806496462247e-06,
+      "loss": 1.2441,
+      "step": 18192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.094838215446545,
+      "learning_rate": 2.5499711704181835e-06,
+      "loss": 0.5947,
+      "step": 18193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.169289515579847,
+      "learning_rate": 2.5481364691492814e-06,
+      "loss": 2.0195,
+      "step": 18194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.089881511596633,
+      "learning_rate": 2.546302392706598e-06,
+      "loss": 1.3867,
+      "step": 18195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.625724576270744,
+      "learning_rate": 2.5444689411411786e-06,
+      "loss": 1.2979,
+      "step": 18196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0473490028431915,
+      "learning_rate": 2.542636114504052e-06,
+      "loss": 1.2617,
+      "step": 18197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3161152597383454,
+      "learning_rate": 2.5408039128462207e-06,
+      "loss": 1.4805,
+      "step": 18198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.356715237855713,
+      "learning_rate": 2.5389723362186647e-06,
+      "loss": 1.0234,
+      "step": 18199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.48708461018401,
+      "learning_rate": 2.537141384672373e-06,
+      "loss": 0.8467,
+      "step": 18200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6961217961393182,
+      "learning_rate": 2.535311058258291e-06,
+      "loss": 0.921,
+      "step": 18201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.617751149076476,
+      "learning_rate": 2.533481357027356e-06,
+      "loss": 1.1074,
+      "step": 18202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.81123919041628,
+      "learning_rate": 2.531652281030486e-06,
+      "loss": 0.9883,
+      "step": 18203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1108004804369576,
+      "learning_rate": 2.529823830318584e-06,
+      "loss": 0.9707,
+      "step": 18204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.788162848254095,
+      "learning_rate": 2.5279960049425428e-06,
+      "loss": 0.9932,
+      "step": 18205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.182099233010974,
+      "learning_rate": 2.5261688049532246e-06,
+      "loss": 0.9868,
+      "step": 18206
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9372297122055646,
+      "learning_rate": 2.5243422304014803e-06,
+      "loss": 0.9297,
+      "step": 18207
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.103373158744794,
+      "learning_rate": 2.522516281338136e-06,
+      "loss": 0.8975,
+      "step": 18208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6345220160499245,
+      "learning_rate": 2.5206909578140177e-06,
+      "loss": 1.1602,
+      "step": 18209
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6941287703362713,
+      "learning_rate": 2.518866259879918e-06,
+      "loss": 1.2529,
+      "step": 18210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8724129688700595,
+      "learning_rate": 2.5170421875866186e-06,
+      "loss": 1.0635,
+      "step": 18211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9645913418889436,
+      "learning_rate": 2.515218740984879e-06,
+      "loss": 1.6699,
+      "step": 18212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4839364870401113,
+      "learning_rate": 2.5133959201254527e-06,
+      "loss": 0.6733,
+      "step": 18213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.695757655616484,
+      "learning_rate": 2.5115737250590578e-06,
+      "loss": 1.0508,
+      "step": 18214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8444308786636747,
+      "learning_rate": 2.5097521558364144e-06,
+      "loss": 1.1436,
+      "step": 18215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8373255189873943,
+      "learning_rate": 2.5079312125082163e-06,
+      "loss": 1.4023,
+      "step": 18216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9672138942556714,
+      "learning_rate": 2.50611089512513e-06,
+      "loss": 0.6357,
+      "step": 18217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.805936967064468,
+      "learning_rate": 2.504291203737824e-06,
+      "loss": 0.6233,
+      "step": 18218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2147988612555976,
+      "learning_rate": 2.502472138396936e-06,
+      "loss": 1.7324,
+      "step": 18219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.332944655970704,
+      "learning_rate": 2.5006536991530854e-06,
+      "loss": 0.9238,
+      "step": 18220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3923745980976747,
+      "learning_rate": 2.4988358860568885e-06,
+      "loss": 0.8359,
+      "step": 18221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2788936328440683,
+      "learning_rate": 2.497018699158926e-06,
+      "loss": 1.2783,
+      "step": 18222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7748492583436724,
+      "learning_rate": 2.4952021385097696e-06,
+      "loss": 1.3828,
+      "step": 18223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9264286338167227,
+      "learning_rate": 2.493386204159981e-06,
+      "loss": 0.8984,
+      "step": 18224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.03032287566822,
+      "learning_rate": 2.4915708961600866e-06,
+      "loss": 0.8872,
+      "step": 18225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.037307036122171,
+      "learning_rate": 2.489756214560615e-06,
+      "loss": 0.9229,
+      "step": 18226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2768511231160278,
+      "learning_rate": 2.487942159412063e-06,
+      "loss": 0.6504,
+      "step": 18227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4675884879660965,
+      "learning_rate": 2.4861287307649165e-06,
+      "loss": 1.2598,
+      "step": 18228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8912595310210762,
+      "learning_rate": 2.4843159286696395e-06,
+      "loss": 1.1465,
+      "step": 18229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1232670365838917,
+      "learning_rate": 2.4825037531766876e-06,
+      "loss": 0.875,
+      "step": 18230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.54991950861733,
+      "learning_rate": 2.4806922043364905e-06,
+      "loss": 0.6266,
+      "step": 18231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6762914908986635,
+      "learning_rate": 2.478881282199455e-06,
+      "loss": 0.75,
+      "step": 18232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8974778375349097,
+      "learning_rate": 2.477070986815991e-06,
+      "loss": 1.0439,
+      "step": 18233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.57048010931832,
+      "learning_rate": 2.475261318236466e-06,
+      "loss": 1.2617,
+      "step": 18234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6464193300931957,
+      "learning_rate": 2.4734522765112566e-06,
+      "loss": 0.9917,
+      "step": 18235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.481480369815962,
+      "learning_rate": 2.4716438616906977e-06,
+      "loss": 1.5195,
+      "step": 18236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8722781452514035,
+      "learning_rate": 2.469836073825113e-06,
+      "loss": 0.6953,
+      "step": 18237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.541062486762448,
+      "learning_rate": 2.4680289129648232e-06,
+      "loss": 1.208,
+      "step": 18238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9153671679051096,
+      "learning_rate": 2.4662223791601186e-06,
+      "loss": 0.7236,
+      "step": 18239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6205145342284943,
+      "learning_rate": 2.46441647246127e-06,
+      "loss": 0.7939,
+      "step": 18240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.092536105659063,
+      "learning_rate": 2.4626111929185317e-06,
+      "loss": 0.9599,
+      "step": 18241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.823476031521977,
+      "learning_rate": 2.4608065405821503e-06,
+      "loss": 0.9785,
+      "step": 18242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.917851623158464,
+      "learning_rate": 2.459002515502351e-06,
+      "loss": 1.2646,
+      "step": 18243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6989544574842714,
+      "learning_rate": 2.4571991177293335e-06,
+      "loss": 1.2012,
+      "step": 18244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8067169072901628,
+      "learning_rate": 2.4553963473132906e-06,
+      "loss": 0.8999,
+      "step": 18245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.29036977099359,
+      "learning_rate": 2.4535942043043824e-06,
+      "loss": 1.5391,
+      "step": 18246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.491160081217554,
+      "learning_rate": 2.4517926887527744e-06,
+      "loss": 0.8054,
+      "step": 18247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6229907020581478,
+      "learning_rate": 2.4499918007085983e-06,
+      "loss": 0.4872,
+      "step": 18248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0642397862821493,
+      "learning_rate": 2.448191540221967e-06,
+      "loss": 1.4961,
+      "step": 18249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.995976812292964,
+      "learning_rate": 2.446391907342982e-06,
+      "loss": 0.9495,
+      "step": 18250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.657350325627613,
+      "learning_rate": 2.4445929021217285e-06,
+      "loss": 1.9551,
+      "step": 18251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.004565409387697,
+      "learning_rate": 2.4427945246082744e-06,
+      "loss": 0.832,
+      "step": 18252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.163283866868052,
+      "learning_rate": 2.440996774852669e-06,
+      "loss": 0.9561,
+      "step": 18253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.001388425501213,
+      "learning_rate": 2.439199652904936e-06,
+      "loss": 0.6816,
+      "step": 18254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.804465410772922,
+      "learning_rate": 2.437403158815091e-06,
+      "loss": 1.0625,
+      "step": 18255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.540234104692051,
+      "learning_rate": 2.4356072926331324e-06,
+      "loss": 1.459,
+      "step": 18256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.306185211756639,
+      "learning_rate": 2.4338120544090405e-06,
+      "loss": 0.542,
+      "step": 18257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.403000616947867,
+      "learning_rate": 2.4320174441927635e-06,
+      "loss": 1.3281,
+      "step": 18258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.029871741968207,
+      "learning_rate": 2.430223462034262e-06,
+      "loss": 0.7598,
+      "step": 18259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7774401292764355,
+      "learning_rate": 2.4284301079834454e-06,
+      "loss": 0.5403,
+      "step": 18260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5370814692524957,
+      "learning_rate": 2.4266373820902383e-06,
+      "loss": 0.9902,
+      "step": 18261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.532278346522668,
+      "learning_rate": 2.42484528440452e-06,
+      "loss": 1.1758,
+      "step": 18262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.730912267074166,
+      "learning_rate": 2.4230538149761645e-06,
+      "loss": 0.7407,
+      "step": 18263
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2905409273172954,
+      "learning_rate": 2.4212629738550347e-06,
+      "loss": 1.0518,
+      "step": 18264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.993269960593853,
+      "learning_rate": 2.419472761090963e-06,
+      "loss": 0.9062,
+      "step": 18265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.9798565417236125,
+      "learning_rate": 2.4176831767337733e-06,
+      "loss": 1.0586,
+      "step": 18266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.648840049375903,
+      "learning_rate": 2.4158942208332617e-06,
+      "loss": 0.906,
+      "step": 18267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.480477624669846,
+      "learning_rate": 2.414105893439225e-06,
+      "loss": 1.2422,
+      "step": 18268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2423954777509016,
+      "learning_rate": 2.4123181946014234e-06,
+      "loss": 0.9626,
+      "step": 18269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7273415444950637,
+      "learning_rate": 2.4105311243696137e-06,
+      "loss": 1.0938,
+      "step": 18270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.61873378338948,
+      "learning_rate": 2.408744682793529e-06,
+      "loss": 1.5352,
+      "step": 18271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6449489967894797,
+      "learning_rate": 2.406958869922876e-06,
+      "loss": 0.5146,
+      "step": 18272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.485246997124984,
+      "learning_rate": 2.405173685807366e-06,
+      "loss": 0.8135,
+      "step": 18273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1286764262895455,
+      "learning_rate": 2.403389130496672e-06,
+      "loss": 0.918,
+      "step": 18274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0560056374183304,
+      "learning_rate": 2.401605204040461e-06,
+      "loss": 1.0596,
+      "step": 18275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.250252815454182,
+      "learning_rate": 2.3998219064883724e-06,
+      "loss": 1.3184,
+      "step": 18276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4753244476357503,
+      "learning_rate": 2.398039237890046e-06,
+      "loss": 1.001,
+      "step": 18277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.765706637626959,
+      "learning_rate": 2.396257198295082e-06,
+      "loss": 0.7749,
+      "step": 18278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6368646807539764,
+      "learning_rate": 2.3944757877530784e-06,
+      "loss": 1.377,
+      "step": 18279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3679549544124208,
+      "learning_rate": 2.3926950063136083e-06,
+      "loss": 0.6987,
+      "step": 18280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.6093958967349415,
+      "learning_rate": 2.3909148540262383e-06,
+      "loss": 0.8884,
+      "step": 18281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.526115127466895,
+      "learning_rate": 2.3891353309405007e-06,
+      "loss": 0.7795,
+      "step": 18282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.143546011870257,
+      "learning_rate": 2.3873564371059253e-06,
+      "loss": 0.7852,
+      "step": 18283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.833395583273016,
+      "learning_rate": 2.385578172572009e-06,
+      "loss": 0.6953,
+      "step": 18284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9062093511531162,
+      "learning_rate": 2.3838005373882503e-06,
+      "loss": 1.3125,
+      "step": 18285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6231011380102585,
+      "learning_rate": 2.3820235316041163e-06,
+      "loss": 0.6479,
+      "step": 18286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3790814314860267,
+      "learning_rate": 2.380247155269061e-06,
+      "loss": 1.001,
+      "step": 18287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1909656863779263,
+      "learning_rate": 2.3784714084325118e-06,
+      "loss": 0.9189,
+      "step": 18288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.528259342746493,
+      "learning_rate": 2.376696291143893e-06,
+      "loss": 1.041,
+      "step": 18289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1317464658560925,
+      "learning_rate": 2.374921803452615e-06,
+      "loss": 0.792,
+      "step": 18290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.473994508904755,
+      "learning_rate": 2.3731479454080496e-06,
+      "loss": 0.9854,
+      "step": 18291
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0540919198902903,
+      "learning_rate": 2.371374717059566e-06,
+      "loss": 0.8672,
+      "step": 18292
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.354716470257928,
+      "learning_rate": 2.369602118456507e-06,
+      "loss": 0.9443,
+      "step": 18293
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5612370444129415,
+      "learning_rate": 2.367830149648215e-06,
+      "loss": 0.7231,
+      "step": 18294
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1635051794622675,
+      "learning_rate": 2.366058810683994e-06,
+      "loss": 0.7637,
+      "step": 18295
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.33357558959563,
+      "learning_rate": 2.3642881016131433e-06,
+      "loss": 0.787,
+      "step": 18296
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.455082406289043,
+      "learning_rate": 2.3625180224849304e-06,
+      "loss": 0.9111,
+      "step": 18297
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.7001497610053775,
+      "learning_rate": 2.360748573348634e-06,
+      "loss": 0.5493,
+      "step": 18298
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.574194565211471,
+      "learning_rate": 2.3589797542534876e-06,
+      "loss": 1.5801,
+      "step": 18299
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3865213515443253,
+      "learning_rate": 2.357211565248718e-06,
+      "loss": 0.999,
+      "step": 18300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.086620149788481,
+      "learning_rate": 2.355444006383528e-06,
+      "loss": 1.0156,
+      "step": 18301
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.089043167509134,
+      "learning_rate": 2.3536770777071173e-06,
+      "loss": 1.1943,
+      "step": 18302
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8546353440304237,
+      "learning_rate": 2.351910779268654e-06,
+      "loss": 1.0586,
+      "step": 18303
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.477758365558862,
+      "learning_rate": 2.350145111117294e-06,
+      "loss": 1.1348,
+      "step": 18304
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5813734476006824,
+      "learning_rate": 2.3483800733021725e-06,
+      "loss": 0.7686,
+      "step": 18305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.96696100241514,
+      "learning_rate": 2.3466156658724093e-06,
+      "loss": 0.7524,
+      "step": 18306
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.256575618505547,
+      "learning_rate": 2.3448518888771142e-06,
+      "loss": 1.6104,
+      "step": 18307
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.21036912477325,
+      "learning_rate": 2.3430887423653707e-06,
+      "loss": 0.915,
+      "step": 18308
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.374942843975423,
+      "learning_rate": 2.3413262263862423e-06,
+      "loss": 0.6841,
+      "step": 18309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.093298336676014,
+      "learning_rate": 2.3395643409887757e-06,
+      "loss": 0.8872,
+      "step": 18310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.573006374839005,
+      "learning_rate": 2.3378030862220123e-06,
+      "loss": 0.6489,
+      "step": 18311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.831867434192636,
+      "learning_rate": 2.3360424621349656e-06,
+      "loss": 1.623,
+      "step": 18312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.861826586751769,
+      "learning_rate": 2.3342824687766273e-06,
+      "loss": 0.994,
+      "step": 18313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7068157247863107,
+      "learning_rate": 2.33252310619598e-06,
+      "loss": 1.0381,
+      "step": 18314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3275412463949956,
+      "learning_rate": 2.3307643744419876e-06,
+      "loss": 1.2012,
+      "step": 18315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.590661245690218,
+      "learning_rate": 2.3290062735635914e-06,
+      "loss": 0.5623,
+      "step": 18316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5242081637501435,
+      "learning_rate": 2.3272488036097274e-06,
+      "loss": 1.1484,
+      "step": 18317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4543431653045977,
+      "learning_rate": 2.325491964629292e-06,
+      "loss": 1.2324,
+      "step": 18318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.622556184615934,
+      "learning_rate": 2.3237357566711914e-06,
+      "loss": 1.2129,
+      "step": 18319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4168524067850856,
+      "learning_rate": 2.3219801797842917e-06,
+      "loss": 0.6182,
+      "step": 18320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.686600376048837,
+      "learning_rate": 2.3202252340174507e-06,
+      "loss": 1.1143,
+      "step": 18321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4933179011501507,
+      "learning_rate": 2.318470919419505e-06,
+      "loss": 1.0479,
+      "step": 18322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.711587725359948,
+      "learning_rate": 2.316717236039284e-06,
+      "loss": 1.0586,
+      "step": 18323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.043639044335275,
+      "learning_rate": 2.314964183925586e-06,
+      "loss": 0.9365,
+      "step": 18324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9221232804368047,
+      "learning_rate": 2.3132117631271987e-06,
+      "loss": 0.7666,
+      "step": 18325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.855306117801286,
+      "learning_rate": 2.3114599736928947e-06,
+      "loss": 1.5449,
+      "step": 18326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.935923538865073,
+      "learning_rate": 2.3097088156714184e-06,
+      "loss": 1.248,
+      "step": 18327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8439443333217187,
+      "learning_rate": 2.3079582891115144e-06,
+      "loss": 1.0801,
+      "step": 18328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.778244084831337,
+      "learning_rate": 2.306208394061893e-06,
+      "loss": 1.0605,
+      "step": 18329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.145272379665258,
+      "learning_rate": 2.3044591305712516e-06,
+      "loss": 0.5664,
+      "step": 18330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.050766368993629,
+      "learning_rate": 2.3027104986882708e-06,
+      "loss": 1.1699,
+      "step": 18331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5472246175810427,
+      "learning_rate": 2.30096249846162e-06,
+      "loss": 0.875,
+      "step": 18332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.2898522155489225,
+      "learning_rate": 2.299215129939944e-06,
+      "loss": 1.4941,
+      "step": 18333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.4155381158378635,
+      "learning_rate": 2.2974683931718643e-06,
+      "loss": 1.0586,
+      "step": 18334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2932443794290567,
+      "learning_rate": 2.2957222882059953e-06,
+      "loss": 0.8408,
+      "step": 18335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.971388947465578,
+      "learning_rate": 2.293976815090937e-06,
+      "loss": 0.9038,
+      "step": 18336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.41839799202026,
+      "learning_rate": 2.2922319738752613e-06,
+      "loss": 0.6582,
+      "step": 18337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1439066999676895,
+      "learning_rate": 2.290487764607524e-06,
+      "loss": 1.1621,
+      "step": 18338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.545019564787572,
+      "learning_rate": 2.288744187336264e-06,
+      "loss": 1.292,
+      "step": 18339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.177805864332482,
+      "learning_rate": 2.287001242110012e-06,
+      "loss": 1.5566,
+      "step": 18340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.630907405196777,
+      "learning_rate": 2.2852589289772685e-06,
+      "loss": 0.9941,
+      "step": 18341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6014377452363473,
+      "learning_rate": 2.2835172479865217e-06,
+      "loss": 0.6243,
+      "step": 18342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7226533602421448,
+      "learning_rate": 2.281776199186239e-06,
+      "loss": 0.9131,
+      "step": 18343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6377715996928157,
+      "learning_rate": 2.280035782624876e-06,
+      "loss": 1.1719,
+      "step": 18344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3437679035774517,
+      "learning_rate": 2.278295998350871e-06,
+      "loss": 1.1504,
+      "step": 18345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0116147271979665,
+      "learning_rate": 2.276556846412639e-06,
+      "loss": 0.915,
+      "step": 18346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8893488127599065,
+      "learning_rate": 2.274818326858577e-06,
+      "loss": 0.6499,
+      "step": 18347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4063066950288174,
+      "learning_rate": 2.2730804397370688e-06,
+      "loss": 0.8047,
+      "step": 18348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.246780120961324,
+      "learning_rate": 2.2713431850964805e-06,
+      "loss": 0.4912,
+      "step": 18349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.225278874510274,
+      "learning_rate": 2.2696065629851604e-06,
+      "loss": 1.0127,
+      "step": 18350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7371629868390668,
+      "learning_rate": 2.267870573451436e-06,
+      "loss": 0.2865,
+      "step": 18351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.896272173434067,
+      "learning_rate": 2.266135216543616e-06,
+      "loss": 0.5093,
+      "step": 18352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.923022949486126,
+      "learning_rate": 2.2644004923099983e-06,
+      "loss": 1.0537,
+      "step": 18353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.312163851255901,
+      "learning_rate": 2.26266640079886e-06,
+      "loss": 0.7148,
+      "step": 18354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1330099067421777,
+      "learning_rate": 2.2609329420584636e-06,
+      "loss": 0.8555,
+      "step": 18355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6603426728054904,
+      "learning_rate": 2.2592001161370392e-06,
+      "loss": 0.4927,
+      "step": 18356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.745154620403575,
+      "learning_rate": 2.257467923082823e-06,
+      "loss": 1.5059,
+      "step": 18357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.646175532601876,
+      "learning_rate": 2.2557363629440164e-06,
+      "loss": 0.7725,
+      "step": 18358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3588636235705724,
+      "learning_rate": 2.254005435768808e-06,
+      "loss": 0.4656,
+      "step": 18359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5353235026043097,
+      "learning_rate": 2.2522751416053617e-06,
+      "loss": 0.6924,
+      "step": 18360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0625770714851606,
+      "learning_rate": 2.250545480501845e-06,
+      "loss": 0.8096,
+      "step": 18361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.124943564170452,
+      "learning_rate": 2.2488164525063804e-06,
+      "loss": 0.8689,
+      "step": 18362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.431925247095665,
+      "learning_rate": 2.247088057667096e-06,
+      "loss": 1.291,
+      "step": 18363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.622745417062896,
+      "learning_rate": 2.2453602960320895e-06,
+      "loss": 0.9443,
+      "step": 18364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.660649773390757,
+      "learning_rate": 2.2436331676494364e-06,
+      "loss": 1.2676,
+      "step": 18365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.630419674467352,
+      "learning_rate": 2.241906672567215e-06,
+      "loss": 0.7834,
+      "step": 18366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.84160520858545,
+      "learning_rate": 2.240180810833464e-06,
+      "loss": 1.2612,
+      "step": 18367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.030579911257121,
+      "learning_rate": 2.2384555824962155e-06,
+      "loss": 0.9521,
+      "step": 18368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5216132499683157,
+      "learning_rate": 2.2367309876034775e-06,
+      "loss": 1.0605,
+      "step": 18369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5033994935583308,
+      "learning_rate": 2.2350070262032537e-06,
+      "loss": 0.6033,
+      "step": 18370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.410442613502147,
+      "learning_rate": 2.2332836983435136e-06,
+      "loss": 0.9551,
+      "step": 18371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.217543905483678,
+      "learning_rate": 2.2315610040722247e-06,
+      "loss": 0.7163,
+      "step": 18372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.967285758867904,
+      "learning_rate": 2.2298389434373207e-06,
+      "loss": 0.9133,
+      "step": 18373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.250296573166704,
+      "learning_rate": 2.2281175164867275e-06,
+      "loss": 1.3906,
+      "step": 18374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0429869272964787,
+      "learning_rate": 2.226396723268356e-06,
+      "loss": 0.957,
+      "step": 18375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.965455016573048,
+      "learning_rate": 2.2246765638300943e-06,
+      "loss": 0.7437,
+      "step": 18376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.173766676087832,
+      "learning_rate": 2.2229570382198084e-06,
+      "loss": 0.6987,
+      "step": 18377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0062177121020612,
+      "learning_rate": 2.2212381464853576e-06,
+      "loss": 0.8726,
+      "step": 18378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.508446121052592,
+      "learning_rate": 2.219519888674576e-06,
+      "loss": 1.5439,
+      "step": 18379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.070400020260936,
+      "learning_rate": 2.2178022648352837e-06,
+      "loss": 1.1089,
+      "step": 18380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3493939833717934,
+      "learning_rate": 2.2160852750152727e-06,
+      "loss": 0.708,
+      "step": 18381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0577821814780815,
+      "learning_rate": 2.2143689192623358e-06,
+      "loss": 1.0854,
+      "step": 18382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.994305968651327,
+      "learning_rate": 2.21265319762424e-06,
+      "loss": 0.7019,
+      "step": 18383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.5734226697049865,
+      "learning_rate": 2.2109381101487274e-06,
+      "loss": 1.291,
+      "step": 18384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5149608747711367,
+      "learning_rate": 2.20922365688353e-06,
+      "loss": 0.5356,
+      "step": 18385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.417622908193318,
+      "learning_rate": 2.2075098378763564e-06,
+      "loss": 1.2666,
+      "step": 18386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.856831904928958,
+      "learning_rate": 2.2057966531749103e-06,
+      "loss": 0.9551,
+      "step": 18387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9986441904275531,
+      "learning_rate": 2.2040841028268646e-06,
+      "loss": 1.0,
+      "step": 18388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.429306932677865,
+      "learning_rate": 2.2023721868798754e-06,
+      "loss": 1.0439,
+      "step": 18389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.550686470829597,
+      "learning_rate": 2.2006609053815853e-06,
+      "loss": 0.9165,
+      "step": 18390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.797065435074353,
+      "learning_rate": 2.1989502583796175e-06,
+      "loss": 1.6064,
+      "step": 18391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.735712240387229,
+      "learning_rate": 2.197240245921589e-06,
+      "loss": 0.5903,
+      "step": 18392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8631210371070264,
+      "learning_rate": 2.195530868055079e-06,
+      "loss": 1.2021,
+      "step": 18393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.140395923201455,
+      "learning_rate": 2.1938221248276603e-06,
+      "loss": 2.3574,
+      "step": 18394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.124002671414292,
+      "learning_rate": 2.192114016286889e-06,
+      "loss": 1.1387,
+      "step": 18395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.896892005681524,
+      "learning_rate": 2.1904065424803e-06,
+      "loss": 0.6719,
+      "step": 18396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5049402248710937,
+      "learning_rate": 2.188699703455413e-06,
+      "loss": 1.1328,
+      "step": 18397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9734119953418783,
+      "learning_rate": 2.186993499259721e-06,
+      "loss": 1.2266,
+      "step": 18398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.2918474227221495,
+      "learning_rate": 2.1852879299407104e-06,
+      "loss": 1.2021,
+      "step": 18399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2550651968858486,
+      "learning_rate": 2.1835829955458577e-06,
+      "loss": 0.6806,
+      "step": 18400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.586394209162244,
+      "learning_rate": 2.1818786961225996e-06,
+      "loss": 0.9836,
+      "step": 18401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3611811550888206,
+      "learning_rate": 2.180175031718368e-06,
+      "loss": 0.6439,
+      "step": 18402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.127217708109403,
+      "learning_rate": 2.1784720023805688e-06,
+      "loss": 0.2538,
+      "step": 18403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9584862940145313,
+      "learning_rate": 2.176769608156609e-06,
+      "loss": 0.7488,
+      "step": 18404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.566816158073354,
+      "learning_rate": 2.175067849093862e-06,
+      "loss": 0.7871,
+      "step": 18405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6104075507522135,
+      "learning_rate": 2.1733667252396812e-06,
+      "loss": 0.9238,
+      "step": 18406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.316818507603474,
+      "learning_rate": 2.171666236641409e-06,
+      "loss": 1.333,
+      "step": 18407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.567893582019758,
+      "learning_rate": 2.169966383346372e-06,
+      "loss": 1.1455,
+      "step": 18408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7356633230326577,
+      "learning_rate": 2.1682671654018827e-06,
+      "loss": 0.4333,
+      "step": 18409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8870356512780397,
+      "learning_rate": 2.166568582855222e-06,
+      "loss": 0.4714,
+      "step": 18410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.874963083398999,
+      "learning_rate": 2.1648706357536604e-06,
+      "loss": 1.0518,
+      "step": 18411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.458053691497984,
+      "learning_rate": 2.163173324144449e-06,
+      "loss": 0.7852,
+      "step": 18412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.437481708947047,
+      "learning_rate": 2.161476648074834e-06,
+      "loss": 1.0967,
+      "step": 18413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8006303769540795,
+      "learning_rate": 2.159780607592024e-06,
+      "loss": 0.8076,
+      "step": 18414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9870764204631612,
+      "learning_rate": 2.158085202743218e-06,
+      "loss": 0.9639,
+      "step": 18415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7794177447564437,
+      "learning_rate": 2.1563904335756056e-06,
+      "loss": 1.2256,
+      "step": 18416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6289422541472587,
+      "learning_rate": 2.1546963001363495e-06,
+      "loss": 1.0625,
+      "step": 18417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6967665242460803,
+      "learning_rate": 2.153002802472587e-06,
+      "loss": 1.0854,
+      "step": 18418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.412211180480648,
+      "learning_rate": 2.1513099406314628e-06,
+      "loss": 1.1318,
+      "step": 18419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7254962749214307,
+      "learning_rate": 2.1496177146600765e-06,
+      "loss": 0.9131,
+      "step": 18420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5163347652766594,
+      "learning_rate": 2.1479261246055317e-06,
+      "loss": 1.1182,
+      "step": 18421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.556517394833029,
+      "learning_rate": 2.146235170514899e-06,
+      "loss": 1.2666,
+      "step": 18422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9265272668936593,
+      "learning_rate": 2.1445448524352373e-06,
+      "loss": 1.3027,
+      "step": 18423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.981679873356981,
+      "learning_rate": 2.1428551704135825e-06,
+      "loss": 0.5314,
+      "step": 18424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7080360789765452,
+      "learning_rate": 2.141166124496971e-06,
+      "loss": 1.171,
+      "step": 18425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3833671549640285,
+      "learning_rate": 2.139477714732396e-06,
+      "loss": 0.8076,
+      "step": 18426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.950173492502476,
+      "learning_rate": 2.1377899411668485e-06,
+      "loss": 1.6426,
+      "step": 18427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.621632055150975,
+      "learning_rate": 2.1361028038473034e-06,
+      "loss": 0.8354,
+      "step": 18428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7528876312940835,
+      "learning_rate": 2.134416302820705e-06,
+      "loss": 0.4248,
+      "step": 18429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1829209493949553,
+      "learning_rate": 2.132730438133998e-06,
+      "loss": 1.1719,
+      "step": 18430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.368958267275877,
+      "learning_rate": 2.13104520983409e-06,
+      "loss": 1.1357,
+      "step": 18431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.292553979693103,
+      "learning_rate": 2.1293606179678877e-06,
+      "loss": 0.9863,
+      "step": 18432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.890499627771042,
+      "learning_rate": 2.127676662582262e-06,
+      "loss": 2.2031,
+      "step": 18433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.643655960407919,
+      "learning_rate": 2.1259933437240892e-06,
+      "loss": 1.5332,
+      "step": 18434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3505506433036176,
+      "learning_rate": 2.1243106614402095e-06,
+      "loss": 0.6057,
+      "step": 18435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.083655641223167,
+      "learning_rate": 2.1226286157774445e-06,
+      "loss": 1.2012,
+      "step": 18436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.222300941509213,
+      "learning_rate": 2.120947206782614e-06,
+      "loss": 1.5723,
+      "step": 18437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.810701504822957,
+      "learning_rate": 2.1192664345025148e-06,
+      "loss": 0.9072,
+      "step": 18438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.930913480462372,
+      "learning_rate": 2.1175862989839117e-06,
+      "loss": 0.637,
+      "step": 18439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2088014500605433,
+      "learning_rate": 2.1159068002735705e-06,
+      "loss": 1.1914,
+      "step": 18440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.200461071682689,
+      "learning_rate": 2.11422793841822e-06,
+      "loss": 0.7861,
+      "step": 18441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2114758594289827,
+      "learning_rate": 2.1125497134645954e-06,
+      "loss": 0.6028,
+      "step": 18442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.038693593195469,
+      "learning_rate": 2.1108721254593925e-06,
+      "loss": 0.9824,
+      "step": 18443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.049646613488698,
+      "learning_rate": 2.109195174449302e-06,
+      "loss": 1.1973,
+      "step": 18444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.097947713770119,
+      "learning_rate": 2.1075188604809863e-06,
+      "loss": 1.2764,
+      "step": 18445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.777327317381968,
+      "learning_rate": 2.1058431836011e-06,
+      "loss": 0.5601,
+      "step": 18446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9685340036251202,
+      "learning_rate": 2.1041681438562832e-06,
+      "loss": 0.5752,
+      "step": 18447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.844826658630095,
+      "learning_rate": 2.102493741293146e-06,
+      "loss": 0.8945,
+      "step": 18448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6743902670185378,
+      "learning_rate": 2.1008199759582873e-06,
+      "loss": 0.9893,
+      "step": 18449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.071446858806125,
+      "learning_rate": 2.0991468478982786e-06,
+      "loss": 0.825,
+      "step": 18450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2221634088786715,
+      "learning_rate": 2.097474357159698e-06,
+      "loss": 0.9448,
+      "step": 18451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.330080234205912,
+      "learning_rate": 2.095802503789082e-06,
+      "loss": 1.4473,
+      "step": 18452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.21848469077196,
+      "learning_rate": 2.0941312878329566e-06,
+      "loss": 0.8721,
+      "step": 18453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5485964147252007,
+      "learning_rate": 2.092460709337829e-06,
+      "loss": 1.0283,
+      "step": 18454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9289075997441738,
+      "learning_rate": 2.0907907683501926e-06,
+      "loss": 0.6973,
+      "step": 18455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8382741140866663,
+      "learning_rate": 2.08912146491653e-06,
+      "loss": 0.9756,
+      "step": 18456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.007485587449393,
+      "learning_rate": 2.087452799083289e-06,
+      "loss": 0.9141,
+      "step": 18457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2491821286081186,
+      "learning_rate": 2.0857847708969037e-06,
+      "loss": 1.2793,
+      "step": 18458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8609556404842853,
+      "learning_rate": 2.084117380403805e-06,
+      "loss": 1.041,
+      "step": 18459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.016099645390194,
+      "learning_rate": 2.0824506276503897e-06,
+      "loss": 0.8984,
+      "step": 18460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1539735802222957,
+      "learning_rate": 2.0807845126830454e-06,
+      "loss": 0.6133,
+      "step": 18461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.161864654349424,
+      "learning_rate": 2.07911903554813e-06,
+      "loss": 0.9062,
+      "step": 18462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9755119577643097,
+      "learning_rate": 2.0774541962920086e-06,
+      "loss": 0.6628,
+      "step": 18463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.026497430867059,
+      "learning_rate": 2.075789994961e-06,
+      "loss": 1.02,
+      "step": 18464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4546194201862175,
+      "learning_rate": 2.074126431601425e-06,
+      "loss": 1.0918,
+      "step": 18465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.539963782345134,
+      "learning_rate": 2.072463506259581e-06,
+      "loss": 1.209,
+      "step": 18466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.236902154937256,
+      "learning_rate": 2.0708012189817384e-06,
+      "loss": 0.6924,
+      "step": 18467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5142397715836315,
+      "learning_rate": 2.0691395698141686e-06,
+      "loss": 0.7148,
+      "step": 18468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.085134511201482,
+      "learning_rate": 2.0674785588031097e-06,
+      "loss": 0.749,
+      "step": 18469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2290759330225707,
+      "learning_rate": 2.065818185994786e-06,
+      "loss": 1.1699,
+      "step": 18470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1657443895774793,
+      "learning_rate": 2.0641584514354023e-06,
+      "loss": 0.7893,
+      "step": 18471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7012033382970515,
+      "learning_rate": 2.0624993551711545e-06,
+      "loss": 0.173,
+      "step": 18472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.58292576019914,
+      "learning_rate": 2.0608408972482063e-06,
+      "loss": 1.1387,
+      "step": 18473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8954855491592473,
+      "learning_rate": 2.0591830777127257e-06,
+      "loss": 0.5122,
+      "step": 18474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3266281333873136,
+      "learning_rate": 2.057525896610832e-06,
+      "loss": 1.2891,
+      "step": 18475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.56690101967788,
+      "learning_rate": 2.0558693539886596e-06,
+      "loss": 1.3418,
+      "step": 18476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4620087729080042,
+      "learning_rate": 2.054213449892303e-06,
+      "loss": 0.9492,
+      "step": 18477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2235302003045203,
+      "learning_rate": 2.052558184367845e-06,
+      "loss": 0.6919,
+      "step": 18478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7442814854452016,
+      "learning_rate": 2.050903557461345e-06,
+      "loss": 0.5198,
+      "step": 18479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7699522109506325,
+      "learning_rate": 2.0492495692188614e-06,
+      "loss": 0.9834,
+      "step": 18480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.967266376157473,
+      "learning_rate": 2.0475962196864178e-06,
+      "loss": 0.8584,
+      "step": 18481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1175916213802273,
+      "learning_rate": 2.0459435089100256e-06,
+      "loss": 0.9697,
+      "step": 18482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.200255102819618,
+      "learning_rate": 2.044291436935683e-06,
+      "loss": 1.1514,
+      "step": 18483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2981884373428514,
+      "learning_rate": 2.0426400038093625e-06,
+      "loss": 0.9561,
+      "step": 18484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.947079610186211,
+      "learning_rate": 2.0409892095770265e-06,
+      "loss": 1.0791,
+      "step": 18485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1875873003296853,
+      "learning_rate": 2.0393390542846167e-06,
+      "loss": 0.792,
+      "step": 18486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.750443162756274,
+      "learning_rate": 2.037689537978052e-06,
+      "loss": 0.8867,
+      "step": 18487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.376669859161324,
+      "learning_rate": 2.036040660703234e-06,
+      "loss": 0.6345,
+      "step": 18488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6271941703658201,
+      "learning_rate": 2.034392422506062e-06,
+      "loss": 0.5735,
+      "step": 18489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.167538564374285,
+      "learning_rate": 2.0327448234324003e-06,
+      "loss": 0.8711,
+      "step": 18490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0803137185254315,
+      "learning_rate": 2.0310978635281004e-06,
+      "loss": 0.8213,
+      "step": 18491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.147842508237811,
+      "learning_rate": 2.0294515428389836e-06,
+      "loss": 0.5188,
+      "step": 18492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.572047864314632,
+      "learning_rate": 2.0278058614108917e-06,
+      "loss": 0.7949,
+      "step": 18493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.365659214109984,
+      "learning_rate": 2.0261608192896103e-06,
+      "loss": 1.2305,
+      "step": 18494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.094242095342943,
+      "learning_rate": 2.024516416520919e-06,
+      "loss": 1.4336,
+      "step": 18495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.304228138807837,
+      "learning_rate": 2.0228726531505787e-06,
+      "loss": 1.0947,
+      "step": 18496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.109388950879169,
+      "learning_rate": 2.0212295292243444e-06,
+      "loss": 1.5586,
+      "step": 18497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.632203549381265,
+      "learning_rate": 2.019587044787935e-06,
+      "loss": 0.9058,
+      "step": 18498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.96253434560887,
+      "learning_rate": 2.017945199887064e-06,
+      "loss": 1.5215,
+      "step": 18499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.216944555441414,
+      "learning_rate": 2.016303994567417e-06,
+      "loss": 1.2783,
+      "step": 18500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0431567857499964,
+      "learning_rate": 2.0146634288746744e-06,
+      "loss": 1.2734,
+      "step": 18501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.468092009646292,
+      "learning_rate": 2.013023502854497e-06,
+      "loss": 0.7168,
+      "step": 18502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.142536852976342,
+      "learning_rate": 2.0113842165525143e-06,
+      "loss": 1.124,
+      "step": 18503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.501883006842782,
+      "learning_rate": 2.009745570014354e-06,
+      "loss": 1.2676,
+      "step": 18504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9738789473770355,
+      "learning_rate": 2.008107563285608e-06,
+      "loss": 1.0293,
+      "step": 18505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.037994184454877,
+      "learning_rate": 2.006470196411872e-06,
+      "loss": 1.3359,
+      "step": 18506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7336189752972306,
+      "learning_rate": 2.0048334694387106e-06,
+      "loss": 0.7175,
+      "step": 18507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.755043935198643,
+      "learning_rate": 2.003197382411673e-06,
+      "loss": 0.8199,
+      "step": 18508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7871602728085123,
+      "learning_rate": 2.0015619353762865e-06,
+      "loss": 0.4775,
+      "step": 18509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4722202874889256,
+      "learning_rate": 1.999927128378068e-06,
+      "loss": 0.7964,
+      "step": 18510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1672747104760495,
+      "learning_rate": 1.9982929614625196e-06,
+      "loss": 1.1396,
+      "step": 18511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7984112830454224,
+      "learning_rate": 1.996659434675113e-06,
+      "loss": 1.2676,
+      "step": 18512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6593005446737035,
+      "learning_rate": 1.9950265480613046e-06,
+      "loss": 0.8749,
+      "step": 18513
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.340579588959181,
+      "learning_rate": 1.9933943016665457e-06,
+      "loss": 1.666,
+      "step": 18514
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5733703233929583,
+      "learning_rate": 1.991762695536259e-06,
+      "loss": 0.5703,
+      "step": 18515
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.552801900377705,
+      "learning_rate": 1.9901317297158496e-06,
+      "loss": 1.418,
+      "step": 18516
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0861363600384646,
+      "learning_rate": 1.9885014042507032e-06,
+      "loss": 0.9883,
+      "step": 18517
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.8590007715638635,
+      "learning_rate": 1.986871719186198e-06,
+      "loss": 1.3887,
+      "step": 18518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.470841778512867,
+      "learning_rate": 1.9852426745676862e-06,
+      "loss": 1.2673,
+      "step": 18519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.106275348064537,
+      "learning_rate": 1.9836142704404955e-06,
+      "loss": 1.1201,
+      "step": 18520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5906197836608276,
+      "learning_rate": 1.9819865068499533e-06,
+      "loss": 0.792,
+      "step": 18521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6314644008100725,
+      "learning_rate": 1.9803593838413513e-06,
+      "loss": 0.8467,
+      "step": 18522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0595215764074295,
+      "learning_rate": 1.9787329014599813e-06,
+      "loss": 0.7906,
+      "step": 18523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.12496252217391,
+      "learning_rate": 1.9771070597511043e-06,
+      "loss": 1.4746,
+      "step": 18524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.434611814054248,
+      "learning_rate": 1.9754818587599643e-06,
+      "loss": 1.1621,
+      "step": 18525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.577859393927101,
+      "learning_rate": 1.9738572985317865e-06,
+      "loss": 1.1836,
+      "step": 18526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0998356621774845,
+      "learning_rate": 1.97223337911179e-06,
+      "loss": 0.7847,
+      "step": 18527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.643158696328096,
+      "learning_rate": 1.9706101005451644e-06,
+      "loss": 1.2363,
+      "step": 18528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1965571360142637,
+      "learning_rate": 1.968987462877081e-06,
+      "loss": 0.6797,
+      "step": 18529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.296030836616018,
+      "learning_rate": 1.9673654661526987e-06,
+      "loss": 0.8076,
+      "step": 18530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.289616264064081,
+      "learning_rate": 1.9657441104171644e-06,
+      "loss": 1.1123,
+      "step": 18531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.737869096475324,
+      "learning_rate": 1.964123395715592e-06,
+      "loss": 1.3281,
+      "step": 18532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.863849544716854,
+      "learning_rate": 1.96250332209309e-06,
+      "loss": 1.7109,
+      "step": 18533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7886627950144205,
+      "learning_rate": 1.960883889594739e-06,
+      "loss": 1.1533,
+      "step": 18534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9999863306369896,
+      "learning_rate": 1.959265098265611e-06,
+      "loss": 1.0417,
+      "step": 18535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1993827690904384,
+      "learning_rate": 1.957646948150757e-06,
+      "loss": 0.6479,
+      "step": 18536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.941995305911429,
+      "learning_rate": 1.9560294392952094e-06,
+      "loss": 0.5996,
+      "step": 18537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.646180307860786,
+      "learning_rate": 1.9544125717439743e-06,
+      "loss": 1.2207,
+      "step": 18538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.828617156497165,
+      "learning_rate": 1.9527963455420574e-06,
+      "loss": 0.915,
+      "step": 18539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6985496869976542,
+      "learning_rate": 1.951180760734439e-06,
+      "loss": 1.2085,
+      "step": 18540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8648800413040005,
+      "learning_rate": 1.9495658173660753e-06,
+      "loss": 0.9453,
+      "step": 18541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.11094878085699,
+      "learning_rate": 1.9479515154819133e-06,
+      "loss": 1.04,
+      "step": 18542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3240620023299354,
+      "learning_rate": 1.946337855126873e-06,
+      "loss": 0.5843,
+      "step": 18543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.58468627039461,
+      "learning_rate": 1.944724836345868e-06,
+      "loss": 1.3066,
+      "step": 18544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.636845809974298,
+      "learning_rate": 1.9431124591837825e-06,
+      "loss": 1.0469,
+      "step": 18545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5624396154220546,
+      "learning_rate": 1.9415007236854937e-06,
+      "loss": 1.5098,
+      "step": 18546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9646752590491634,
+      "learning_rate": 1.939889629895847e-06,
+      "loss": 0.7002,
+      "step": 18547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.74841351328494,
+      "learning_rate": 1.9382791778596867e-06,
+      "loss": 0.9082,
+      "step": 18548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.426310163692394,
+      "learning_rate": 1.9366693676218293e-06,
+      "loss": 1.2109,
+      "step": 18549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2967890509269693,
+      "learning_rate": 1.935060199227076e-06,
+      "loss": 0.9824,
+      "step": 18550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.883504810269046,
+      "learning_rate": 1.9334516727202044e-06,
+      "loss": 0.9121,
+      "step": 18551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.46782097055528,
+      "learning_rate": 1.9318437881459843e-06,
+      "loss": 0.4036,
+      "step": 18552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.613444347839307,
+      "learning_rate": 1.9302365455491633e-06,
+      "loss": 0.8486,
+      "step": 18553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0792665716210115,
+      "learning_rate": 1.928629944974464e-06,
+      "loss": 0.9155,
+      "step": 18554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.100850378156752,
+      "learning_rate": 1.9270239864666003e-06,
+      "loss": 1.2451,
+      "step": 18555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2662391619415265,
+      "learning_rate": 1.925418670070267e-06,
+      "loss": 1.2578,
+      "step": 18556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.378940442621299,
+      "learning_rate": 1.923813995830137e-06,
+      "loss": 0.4136,
+      "step": 18557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9570335037918971,
+      "learning_rate": 1.922209963790872e-06,
+      "loss": 0.9375,
+      "step": 18558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9737270997982828,
+      "learning_rate": 1.920606573997108e-06,
+      "loss": 0.9521,
+      "step": 18559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5572385248745357,
+      "learning_rate": 1.919003826493465e-06,
+      "loss": 0.9941,
+      "step": 18560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7881824711107626,
+      "learning_rate": 1.917401721324552e-06,
+      "loss": 0.6504,
+      "step": 18561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 23.649131762190947,
+      "learning_rate": 1.915800258534953e-06,
+      "loss": 1.4971,
+      "step": 18562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.600709304847666,
+      "learning_rate": 1.914199438169234e-06,
+      "loss": 1.6621,
+      "step": 18563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.756380893192356,
+      "learning_rate": 1.9125992602719437e-06,
+      "loss": 1.2881,
+      "step": 18564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.505198890419741,
+      "learning_rate": 1.9109997248876218e-06,
+      "loss": 0.4351,
+      "step": 18565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.395437199057156,
+      "learning_rate": 1.9094008320607733e-06,
+      "loss": 0.7839,
+      "step": 18566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.263144194399389,
+      "learning_rate": 1.9078025818359018e-06,
+      "loss": 1.0762,
+      "step": 18567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4851075058215737,
+      "learning_rate": 1.906204974257486e-06,
+      "loss": 0.8682,
+      "step": 18568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4709469211942023,
+      "learning_rate": 1.9046080093699792e-06,
+      "loss": 0.5629,
+      "step": 18569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2422718205031935,
+      "learning_rate": 1.9030116872178316e-06,
+      "loss": 1.1699,
+      "step": 18570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9594161443220965,
+      "learning_rate": 1.901416007845469e-06,
+      "loss": 0.9146,
+      "step": 18571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7327962622969837,
+      "learning_rate": 1.8998209712972892e-06,
+      "loss": 0.8965,
+      "step": 18572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.565890163108108,
+      "learning_rate": 1.8982265776176927e-06,
+      "loss": 0.8911,
+      "step": 18573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0894589078705144,
+      "learning_rate": 1.8966328268510469e-06,
+      "loss": 0.8628,
+      "step": 18574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.765077052899392,
+      "learning_rate": 1.895039719041697e-06,
+      "loss": 1.3496,
+      "step": 18575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5028252849616686,
+      "learning_rate": 1.8934472542339931e-06,
+      "loss": 1.2236,
+      "step": 18576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6358527223603767,
+      "learning_rate": 1.8918554324722392e-06,
+      "loss": 1.4473,
+      "step": 18577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0411283635071436,
+      "learning_rate": 1.890264253800747e-06,
+      "loss": 0.9062,
+      "step": 18578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3651038337801933,
+      "learning_rate": 1.888673718263792e-06,
+      "loss": 1.3379,
+      "step": 18579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.277091888461836,
+      "learning_rate": 1.887083825905639e-06,
+      "loss": 1.2656,
+      "step": 18580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.941615260834093,
+      "learning_rate": 1.8854945767705272e-06,
+      "loss": 1.0537,
+      "step": 18581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.5149999161665795,
+      "learning_rate": 1.8839059709026997e-06,
+      "loss": 0.9775,
+      "step": 18582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.022101952979753,
+      "learning_rate": 1.8823180083463565e-06,
+      "loss": 1.1875,
+      "step": 18583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.812539856571142,
+      "learning_rate": 1.8807306891456849e-06,
+      "loss": 0.499,
+      "step": 18584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.228351075418733,
+      "learning_rate": 1.8791440133448746e-06,
+      "loss": 1.0527,
+      "step": 18585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.847337175041526,
+      "learning_rate": 1.8775579809880678e-06,
+      "loss": 0.9961,
+      "step": 18586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.401307709306628,
+      "learning_rate": 1.8759725921194154e-06,
+      "loss": 0.7068,
+      "step": 18587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.825239920849174,
+      "learning_rate": 1.8743878467830294e-06,
+      "loss": 1.2773,
+      "step": 18588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8919846943224026,
+      "learning_rate": 1.8728037450230162e-06,
+      "loss": 0.7832,
+      "step": 18589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.138532186327058,
+      "learning_rate": 1.8712202868834571e-06,
+      "loss": 1.0059,
+      "step": 18590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.947042023182121,
+      "learning_rate": 1.8696374724084226e-06,
+      "loss": 0.6548,
+      "step": 18591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7273554987445314,
+      "learning_rate": 1.8680553016419632e-06,
+      "loss": 0.5864,
+      "step": 18592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8965854537108715,
+      "learning_rate": 1.8664737746281024e-06,
+      "loss": 0.8789,
+      "step": 18593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5559627048180134,
+      "learning_rate": 1.8648928914108604e-06,
+      "loss": 0.9316,
+      "step": 18594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4881289663533206,
+      "learning_rate": 1.8633126520342353e-06,
+      "loss": 1.0212,
+      "step": 18595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.757376873043135,
+      "learning_rate": 1.8617330565421976e-06,
+      "loss": 1.4121,
+      "step": 18596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3627845672108094,
+      "learning_rate": 1.860154104978712e-06,
+      "loss": 1.1074,
+      "step": 18597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.306889342487575,
+      "learning_rate": 1.8585757973877127e-06,
+      "loss": 1.5352,
+      "step": 18598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1108394484262583,
+      "learning_rate": 1.8569981338131314e-06,
+      "loss": 1.041,
+      "step": 18599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.681441213413601,
+      "learning_rate": 1.8554211142988692e-06,
+      "loss": 1.2637,
+      "step": 18600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.4566995346365,
+      "learning_rate": 1.8538447388888186e-06,
+      "loss": 1.7852,
+      "step": 18601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.521937346603617,
+      "learning_rate": 1.852269007626839e-06,
+      "loss": 0.9209,
+      "step": 18602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.317404109119115,
+      "learning_rate": 1.8506939205567898e-06,
+      "loss": 1.2715,
+      "step": 18603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2947085796823363,
+      "learning_rate": 1.8491194777225107e-06,
+      "loss": 1.0391,
+      "step": 18604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.725428216868225,
+      "learning_rate": 1.8475456791678113e-06,
+      "loss": 0.8389,
+      "step": 18605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.549824970927264,
+      "learning_rate": 1.8459725249364873e-06,
+      "loss": 1.1406,
+      "step": 18606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3683411459242487,
+      "learning_rate": 1.8444000150723201e-06,
+      "loss": 1.1768,
+      "step": 18607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.8443408236567205,
+      "learning_rate": 1.8428281496190802e-06,
+      "loss": 1.3164,
+      "step": 18608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9215970491800527,
+      "learning_rate": 1.8412569286205023e-06,
+      "loss": 1.0234,
+      "step": 18609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.029521482848467,
+      "learning_rate": 1.8396863521203178e-06,
+      "loss": 0.4922,
+      "step": 18610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.035025078413126,
+      "learning_rate": 1.838116420162228e-06,
+      "loss": 0.9941,
+      "step": 18611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.448738802375646,
+      "learning_rate": 1.8365471327899315e-06,
+      "loss": 1.0679,
+      "step": 18612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.211859844292734,
+      "learning_rate": 1.8349784900471013e-06,
+      "loss": 0.7402,
+      "step": 18613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7871972595688415,
+      "learning_rate": 1.8334104919773887e-06,
+      "loss": 0.6182,
+      "step": 18614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.423606506590124,
+      "learning_rate": 1.8318431386244257e-06,
+      "loss": 0.6094,
+      "step": 18615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.568287353469795,
+      "learning_rate": 1.8302764300318437e-06,
+      "loss": 1.0859,
+      "step": 18616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.485529505387063,
+      "learning_rate": 1.8287103662432332e-06,
+      "loss": 1.0938,
+      "step": 18617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9451295507462336,
+      "learning_rate": 1.8271449473021813e-06,
+      "loss": 1.4961,
+      "step": 18618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9455781931382852,
+      "learning_rate": 1.8255801732522476e-06,
+      "loss": 0.9756,
+      "step": 18619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8169617759757317,
+      "learning_rate": 1.8240160441369891e-06,
+      "loss": 0.5938,
+      "step": 18620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.49837596715258,
+      "learning_rate": 1.8224525599999238e-06,
+      "loss": 0.3032,
+      "step": 18621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.080296006231147,
+      "learning_rate": 1.8208897208845722e-06,
+      "loss": 1.2969,
+      "step": 18622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4792097321241533,
+      "learning_rate": 1.8193275268344217e-06,
+      "loss": 1.6562,
+      "step": 18623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.350038988203074,
+      "learning_rate": 1.8177659778929463e-06,
+      "loss": 0.8906,
+      "step": 18624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.638942114533057,
+      "learning_rate": 1.8162050741036108e-06,
+      "loss": 1.4395,
+      "step": 18625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.845199323637872,
+      "learning_rate": 1.8146448155098528e-06,
+      "loss": 0.5295,
+      "step": 18626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.299506723333136,
+      "learning_rate": 1.8130852021550877e-06,
+      "loss": 0.6272,
+      "step": 18627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.789049954421271,
+      "learning_rate": 1.8115262340827166e-06,
+      "loss": 1.1172,
+      "step": 18628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.519324952188915,
+      "learning_rate": 1.8099679113361385e-06,
+      "loss": 1.1777,
+      "step": 18629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.362867480870769,
+      "learning_rate": 1.80841023395871e-06,
+      "loss": 1.0488,
+      "step": 18630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.509974798250772,
+      "learning_rate": 1.8068532019937801e-06,
+      "loss": 0.7212,
+      "step": 18631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.625470255555217,
+      "learning_rate": 1.8052968154846834e-06,
+      "loss": 1.063,
+      "step": 18632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.253482737971518,
+      "learning_rate": 1.8037410744747352e-06,
+      "loss": 1.3008,
+      "step": 18633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.128838606293466,
+      "learning_rate": 1.8021859790072342e-06,
+      "loss": 1.5312,
+      "step": 18634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.467320088453113,
+      "learning_rate": 1.8006315291254516e-06,
+      "loss": 0.9775,
+      "step": 18635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.45404835547726,
+      "learning_rate": 1.7990777248726442e-06,
+      "loss": 1.5273,
+      "step": 18636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.328449553842708,
+      "learning_rate": 1.7975245662920636e-06,
+      "loss": 1.1982,
+      "step": 18637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.405429846341647,
+      "learning_rate": 1.795972053426928e-06,
+      "loss": 1.4219,
+      "step": 18638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7879642630596377,
+      "learning_rate": 1.7944201863204447e-06,
+      "loss": 0.9353,
+      "step": 18639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.192094353969078,
+      "learning_rate": 1.7928689650157955e-06,
+      "loss": 0.9312,
+      "step": 18640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.445126851883753,
+      "learning_rate": 1.791318389556157e-06,
+      "loss": 1.0869,
+      "step": 18641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.988999332644589,
+      "learning_rate": 1.7897684599846837e-06,
+      "loss": 0.8379,
+      "step": 18642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6614489670281976,
+      "learning_rate": 1.7882191763445023e-06,
+      "loss": 0.5239,
+      "step": 18643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3270547954433933,
+      "learning_rate": 1.7866705386787335e-06,
+      "loss": 1.0537,
+      "step": 18644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.353686992765257,
+      "learning_rate": 1.7851225470304712e-06,
+      "loss": 1.0273,
+      "step": 18645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.860426025556026,
+      "learning_rate": 1.7835752014427997e-06,
+      "loss": 0.9365,
+      "step": 18646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5758950501914533,
+      "learning_rate": 1.7820285019587797e-06,
+      "loss": 1.0762,
+      "step": 18647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.356658685590655,
+      "learning_rate": 1.7804824486214539e-06,
+      "loss": 1.416,
+      "step": 18648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.002758388922187,
+      "learning_rate": 1.7789370414738383e-06,
+      "loss": 0.5312,
+      "step": 18649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6900692568517517,
+      "learning_rate": 1.7773922805589622e-06,
+      "loss": 0.5486,
+      "step": 18650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.233460886240229,
+      "learning_rate": 1.775848165919805e-06,
+      "loss": 1.1328,
+      "step": 18651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4915196591221176,
+      "learning_rate": 1.7743046975993411e-06,
+      "loss": 1.1719,
+      "step": 18652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.857676905435616,
+      "learning_rate": 1.7727618756405135e-06,
+      "loss": 1.1113,
+      "step": 18653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.485641732354998,
+      "learning_rate": 1.771219700086274e-06,
+      "loss": 1.3574,
+      "step": 18654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6604106394021387,
+      "learning_rate": 1.769678170979533e-06,
+      "loss": 0.8657,
+      "step": 18655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2006699993156187,
+      "learning_rate": 1.7681372883631919e-06,
+      "loss": 0.8887,
+      "step": 18656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.108697035486703,
+      "learning_rate": 1.7665970522801278e-06,
+      "loss": 0.7285,
+      "step": 18657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3989587164125843,
+      "learning_rate": 1.7650574627732148e-06,
+      "loss": 0.4075,
+      "step": 18658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1656318446063714,
+      "learning_rate": 1.7635185198852878e-06,
+      "loss": 0.6257,
+      "step": 18659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.255119733816542,
+      "learning_rate": 1.7619802236591875e-06,
+      "loss": 0.8662,
+      "step": 18660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3046867758539644,
+      "learning_rate": 1.760442574137716e-06,
+      "loss": 1.0664,
+      "step": 18661
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0207289754446394,
+      "learning_rate": 1.7589055713636615e-06,
+      "loss": 0.8628,
+      "step": 18662
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.604503721677216,
+      "learning_rate": 1.7573692153798089e-06,
+      "loss": 1.0156,
+      "step": 18663
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.348510348354855,
+      "learning_rate": 1.7558335062289072e-06,
+      "loss": 1.1426,
+      "step": 18664
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9003490303730413,
+      "learning_rate": 1.7542984439536975e-06,
+      "loss": 1.207,
+      "step": 18665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.29450436095983,
+      "learning_rate": 1.7527640285968956e-06,
+      "loss": 0.9639,
+      "step": 18666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.798046056096262,
+      "learning_rate": 1.7512302602012088e-06,
+      "loss": 0.7153,
+      "step": 18667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.069907435907321,
+      "learning_rate": 1.749697138809317e-06,
+      "loss": 0.6992,
+      "step": 18668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.169005000269162,
+      "learning_rate": 1.7481646644638916e-06,
+      "loss": 0.9639,
+      "step": 18669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.170654942168424,
+      "learning_rate": 1.7466328372075764e-06,
+      "loss": 0.9189,
+      "step": 18670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2587318495408475,
+      "learning_rate": 1.7451016570830037e-06,
+      "loss": 0.8032,
+      "step": 18671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.566640890034257,
+      "learning_rate": 1.7435711241327868e-06,
+      "loss": 1.1494,
+      "step": 18672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9030923799873056,
+      "learning_rate": 1.7420412383995165e-06,
+      "loss": 0.5513,
+      "step": 18673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.624772924865024,
+      "learning_rate": 1.7405119999257673e-06,
+      "loss": 1.1836,
+      "step": 18674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5459498497937454,
+      "learning_rate": 1.7389834087541024e-06,
+      "loss": 1.0215,
+      "step": 18675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0380594994841292,
+      "learning_rate": 1.7374554649270624e-06,
+      "loss": 1.0249,
+      "step": 18676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6466574299104235,
+      "learning_rate": 1.735928168487161e-06,
+      "loss": 0.7863,
+      "step": 18677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4058041840775912,
+      "learning_rate": 1.7344015194769109e-06,
+      "loss": 0.7488,
+      "step": 18678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2809092646057314,
+      "learning_rate": 1.7328755179387924e-06,
+      "loss": 0.9951,
+      "step": 18679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.777370726817854,
+      "learning_rate": 1.7313501639152797e-06,
+      "loss": 1.4453,
+      "step": 18680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3155746639819452,
+      "learning_rate": 1.7298254574488193e-06,
+      "loss": 0.8955,
+      "step": 18681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9520001565059633,
+      "learning_rate": 1.7283013985818441e-06,
+      "loss": 1.0337,
+      "step": 18682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.78811834687271,
+      "learning_rate": 1.7267779873567618e-06,
+      "loss": 1.0352,
+      "step": 18683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1133053931586456,
+      "learning_rate": 1.7252552238159775e-06,
+      "loss": 0.6909,
+      "step": 18684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.524404893384569,
+      "learning_rate": 1.7237331080018653e-06,
+      "loss": 0.875,
+      "step": 18685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6545390284592316,
+      "learning_rate": 1.7222116399567834e-06,
+      "loss": 0.8838,
+      "step": 18686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.059430120769521,
+      "learning_rate": 1.7206908197230726e-06,
+      "loss": 1.2285,
+      "step": 18687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9203043068305674,
+      "learning_rate": 1.719170647343063e-06,
+      "loss": 1.6035,
+      "step": 18688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.037877114623907,
+      "learning_rate": 1.717651122859057e-06,
+      "loss": 0.5186,
+      "step": 18689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6691211519198736,
+      "learning_rate": 1.7161322463133427e-06,
+      "loss": 0.9995,
+      "step": 18690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.739030051499076,
+      "learning_rate": 1.714614017748184e-06,
+      "loss": 1.0439,
+      "step": 18691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9044736950734356,
+      "learning_rate": 1.7130964372058439e-06,
+      "loss": 1.3271,
+      "step": 18692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.75131596681073,
+      "learning_rate": 1.711579504728547e-06,
+      "loss": 0.8916,
+      "step": 18693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.845518712167545,
+      "learning_rate": 1.7100632203585153e-06,
+      "loss": 1.4609,
+      "step": 18694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.734584429758753,
+      "learning_rate": 1.7085475841379373e-06,
+      "loss": 1.1934,
+      "step": 18695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.369532999592313,
+      "learning_rate": 1.7070325961089956e-06,
+      "loss": 1.6904,
+      "step": 18696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.572900367922002,
+      "learning_rate": 1.705518256313862e-06,
+      "loss": 0.7402,
+      "step": 18697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.503565724943391,
+      "learning_rate": 1.7040045647946702e-06,
+      "loss": 1.334,
+      "step": 18698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.29163421839081,
+      "learning_rate": 1.7024915215935465e-06,
+      "loss": 1.0986,
+      "step": 18699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0437716188366446,
+      "learning_rate": 1.700979126752597e-06,
+      "loss": 0.9141,
+      "step": 18700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6939815799292648,
+      "learning_rate": 1.6994673803139155e-06,
+      "loss": 1.416,
+      "step": 18701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5412891190909983,
+      "learning_rate": 1.6979562823195738e-06,
+      "loss": 0.7275,
+      "step": 18702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.130489410796791,
+      "learning_rate": 1.6964458328116217e-06,
+      "loss": 1.011,
+      "step": 18703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.124042516890394,
+      "learning_rate": 1.6949360318320894e-06,
+      "loss": 0.918,
+      "step": 18704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.596462531906815,
+      "learning_rate": 1.6934268794230018e-06,
+      "loss": 1.3184,
+      "step": 18705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.159923023512361,
+      "learning_rate": 1.6919183756263585e-06,
+      "loss": 1.25,
+      "step": 18706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.634376129475959,
+      "learning_rate": 1.6904105204841397e-06,
+      "loss": 0.8682,
+      "step": 18707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.822188918183365,
+      "learning_rate": 1.6889033140383037e-06,
+      "loss": 1.543,
+      "step": 18708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.690956564664999,
+      "learning_rate": 1.6873967563308003e-06,
+      "loss": 0.9395,
+      "step": 18709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1494477827034664,
+      "learning_rate": 1.685890847403554e-06,
+      "loss": 0.894,
+      "step": 18710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7955590247614683,
+      "learning_rate": 1.684385587298476e-06,
+      "loss": 1.1816,
+      "step": 18711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9343834699651752,
+      "learning_rate": 1.6828809760574522e-06,
+      "loss": 0.9844,
+      "step": 18712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3618305546747482,
+      "learning_rate": 1.68137701372236e-06,
+      "loss": 0.6265,
+      "step": 18713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4417810133787832,
+      "learning_rate": 1.6798737003350495e-06,
+      "loss": 0.481,
+      "step": 18714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.646153680392078,
+      "learning_rate": 1.678371035937365e-06,
+      "loss": 0.954,
+      "step": 18715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.296214728917753,
+      "learning_rate": 1.6768690205711173e-06,
+      "loss": 1.1523,
+      "step": 18716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.856879195185174,
+      "learning_rate": 1.6753676542781094e-06,
+      "loss": 1.1699,
+      "step": 18717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.844018073929642,
+      "learning_rate": 1.673866937100127e-06,
+      "loss": 0.8616,
+      "step": 18718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.140771679830045,
+      "learning_rate": 1.6723668690789313e-06,
+      "loss": 1.0889,
+      "step": 18719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2181119404330674,
+      "learning_rate": 1.6708674502562694e-06,
+      "loss": 1.2383,
+      "step": 18720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5638502807851706,
+      "learning_rate": 1.6693686806738661e-06,
+      "loss": 1.3652,
+      "step": 18721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.650463578324213,
+      "learning_rate": 1.6678705603734357e-06,
+      "loss": 1.1514,
+      "step": 18722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.367531520612737,
+      "learning_rate": 1.6663730893966667e-06,
+      "loss": 0.7429,
+      "step": 18723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.919037054277529,
+      "learning_rate": 1.6648762677852397e-06,
+      "loss": 1.668,
+      "step": 18724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5388824888684183,
+      "learning_rate": 1.6633800955808077e-06,
+      "loss": 1.0312,
+      "step": 18725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8040692024795524,
+      "learning_rate": 1.661884572825001e-06,
+      "loss": 1.1128,
+      "step": 18726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1011455363324676,
+      "learning_rate": 1.6603896995594503e-06,
+      "loss": 0.7603,
+      "step": 18727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5792769199485064,
+      "learning_rate": 1.658895475825753e-06,
+      "loss": 0.918,
+      "step": 18728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.960468934162569,
+      "learning_rate": 1.6574019016654897e-06,
+      "loss": 0.8164,
+      "step": 18729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6058741858874463,
+      "learning_rate": 1.6559089771202325e-06,
+      "loss": 0.5986,
+      "step": 18730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.039625884506322,
+      "learning_rate": 1.6544167022315231e-06,
+      "loss": 0.9082,
+      "step": 18731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.438074777766006,
+      "learning_rate": 1.6529250770408927e-06,
+      "loss": 0.9683,
+      "step": 18732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.1097838919967575,
+      "learning_rate": 1.651434101589852e-06,
+      "loss": 1.002,
+      "step": 18733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0184604307184166,
+      "learning_rate": 1.6499437759198932e-06,
+      "loss": 0.9727,
+      "step": 18734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6334657143323925,
+      "learning_rate": 1.6484541000724968e-06,
+      "loss": 0.9863,
+      "step": 18735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.292247990593913,
+      "learning_rate": 1.6469650740891157e-06,
+      "loss": 1.5264,
+      "step": 18736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.211038971314644,
+      "learning_rate": 1.6454766980111918e-06,
+      "loss": 1.1633,
+      "step": 18737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9523283897916108,
+      "learning_rate": 1.6439889718801367e-06,
+      "loss": 1.1523,
+      "step": 18738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.958849216716273,
+      "learning_rate": 1.6425018957373667e-06,
+      "loss": 1.0879,
+      "step": 18739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9749776645797767,
+      "learning_rate": 1.6410154696242603e-06,
+      "loss": 0.4397,
+      "step": 18740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2326783968738284,
+      "learning_rate": 1.6395296935821814e-06,
+      "loss": 1.0464,
+      "step": 18741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6973509226024532,
+      "learning_rate": 1.6380445676524803e-06,
+      "loss": 0.7529,
+      "step": 18742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.77501247123924,
+      "learning_rate": 1.636560091876485e-06,
+      "loss": 1.4453,
+      "step": 18743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.965869572487976,
+      "learning_rate": 1.6350762662955183e-06,
+      "loss": 0.7949,
+      "step": 18744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.953087847991181,
+      "learning_rate": 1.6335930909508663e-06,
+      "loss": 1.3916,
+      "step": 18745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.37997564250589,
+      "learning_rate": 1.6321105658838042e-06,
+      "loss": 1.3623,
+      "step": 18746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.241943452426527,
+      "learning_rate": 1.6306286911355912e-06,
+      "loss": 0.8721,
+      "step": 18747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9943148156735166,
+      "learning_rate": 1.6291474667474687e-06,
+      "loss": 2.0781,
+      "step": 18748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4438823381186663,
+      "learning_rate": 1.6276668927606598e-06,
+      "loss": 1.3594,
+      "step": 18749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8942022509867376,
+      "learning_rate": 1.6261869692163617e-06,
+      "loss": 0.8457,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.604607939121173,
+      "learning_rate": 1.624707696155764e-06,
+      "loss": 0.6978,
+      "step": 18751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7186746149167735,
+      "learning_rate": 1.6232290736200418e-06,
+      "loss": 0.7959,
+      "step": 18752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9121354971203097,
+      "learning_rate": 1.6217511016503346e-06,
+      "loss": 1.106,
+      "step": 18753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.962228503318594,
+      "learning_rate": 1.6202737802877787e-06,
+      "loss": 1.1509,
+      "step": 18754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.609265901946006,
+      "learning_rate": 1.6187971095734833e-06,
+      "loss": 1.0342,
+      "step": 18755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.681958362401215,
+      "learning_rate": 1.6173210895485459e-06,
+      "loss": 1.1895,
+      "step": 18756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5843961394515529,
+      "learning_rate": 1.6158457202540473e-06,
+      "loss": 1.2344,
+      "step": 18757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5038309929121993,
+      "learning_rate": 1.614371001731041e-06,
+      "loss": 0.7183,
+      "step": 18758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.845091996869604,
+      "learning_rate": 1.6128969340205636e-06,
+      "loss": 0.9902,
+      "step": 18759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4322943371466677,
+      "learning_rate": 1.611423517163646e-06,
+      "loss": 1.5098,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6464522063558307,
+      "learning_rate": 1.6099507512012946e-06,
+      "loss": 1.0264,
+      "step": 18761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.810640949851246,
+      "learning_rate": 1.6084786361744902e-06,
+      "loss": 0.8633,
+      "step": 18762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.845740818441481,
+      "learning_rate": 1.607007172124203e-06,
+      "loss": 1.4512,
+      "step": 18763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.356186082210806,
+      "learning_rate": 1.605536359091378e-06,
+      "loss": 0.7725,
+      "step": 18764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1451422927361263,
+      "learning_rate": 1.6040661971169574e-06,
+      "loss": 0.9922,
+      "step": 18765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.484101670302252,
+      "learning_rate": 1.60259668624185e-06,
+      "loss": 1.4375,
+      "step": 18766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2689893275783715,
+      "learning_rate": 1.6011278265069485e-06,
+      "loss": 1.0566,
+      "step": 18767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.59799050652836,
+      "learning_rate": 1.5996596179531364e-06,
+      "loss": 0.7,
+      "step": 18768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7817506216414465,
+      "learning_rate": 1.598192060621273e-06,
+      "loss": 1.3047,
+      "step": 18769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9816069917503625,
+      "learning_rate": 1.596725154552192e-06,
+      "loss": 1.3818,
+      "step": 18770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4735614873254295,
+      "learning_rate": 1.5952588997867279e-06,
+      "loss": 0.9434,
+      "step": 18771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.347428448838902,
+      "learning_rate": 1.5937932963656755e-06,
+      "loss": 0.7268,
+      "step": 18772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.816742084763072,
+      "learning_rate": 1.5923283443298299e-06,
+      "loss": 1.002,
+      "step": 18773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.001581117943437,
+      "learning_rate": 1.5908640437199585e-06,
+      "loss": 0.8496,
+      "step": 18774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.891544708257844,
+      "learning_rate": 1.5894003945768092e-06,
+      "loss": 0.6343,
+      "step": 18775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.155220076221708,
+      "learning_rate": 1.5879373969411137e-06,
+      "loss": 0.8047,
+      "step": 18776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6069511432770742,
+      "learning_rate": 1.5864750508535948e-06,
+      "loss": 1.3438,
+      "step": 18777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.603574507463127,
+      "learning_rate": 1.5850133563549418e-06,
+      "loss": 1.2148,
+      "step": 18778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.122728666754746,
+      "learning_rate": 1.583552313485831e-06,
+      "loss": 1.2256,
+      "step": 18779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1447725002079188,
+      "learning_rate": 1.5820919222869324e-06,
+      "loss": 0.7529,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.587406812823443,
+      "learning_rate": 1.5806321827988774e-06,
+      "loss": 1.207,
+      "step": 18781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.763344741355595,
+      "learning_rate": 1.5791730950622973e-06,
+      "loss": 1.043,
+      "step": 18782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.44335919980041,
+      "learning_rate": 1.577714659117796e-06,
+      "loss": 1.1714,
+      "step": 18783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1066331916676524,
+      "learning_rate": 1.5762568750059604e-06,
+      "loss": 1.0432,
+      "step": 18784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.038893302028583,
+      "learning_rate": 1.574799742767355e-06,
+      "loss": 1.3076,
+      "step": 18785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.614312531529505,
+      "learning_rate": 1.573343262442542e-06,
+      "loss": 0.7507,
+      "step": 18786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5644443857548156,
+      "learning_rate": 1.5718874340720502e-06,
+      "loss": 1.2324,
+      "step": 18787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.398855266159711,
+      "learning_rate": 1.5704322576963859e-06,
+      "loss": 1.0547,
+      "step": 18788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7069352251712124,
+      "learning_rate": 1.5689777333560557e-06,
+      "loss": 0.5957,
+      "step": 18789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1976715557254938,
+      "learning_rate": 1.567523861091541e-06,
+      "loss": 0.8008,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.087665624461931,
+      "learning_rate": 1.5660706409432984e-06,
+      "loss": 1.3887,
+      "step": 18791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5668620124394623,
+      "learning_rate": 1.5646180729517679e-06,
+      "loss": 1.3672,
+      "step": 18792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9925159535719286,
+      "learning_rate": 1.5631661571573724e-06,
+      "loss": 0.5161,
+      "step": 18793
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9524533149717596,
+      "learning_rate": 1.5617148936005272e-06,
+      "loss": 0.6826,
+      "step": 18794
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.908376236963742,
+      "learning_rate": 1.5602642823216135e-06,
+      "loss": 1.625,
+      "step": 18795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.043540969344766,
+      "learning_rate": 1.558814323361002e-06,
+      "loss": 0.8521,
+      "step": 18796
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.00087267565428,
+      "learning_rate": 1.5573650167590409e-06,
+      "loss": 1.4844,
+      "step": 18797
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5341826040144584,
+      "learning_rate": 1.5559163625560702e-06,
+      "loss": 1.3457,
+      "step": 18798
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.451273702383253,
+      "learning_rate": 1.554468360792405e-06,
+      "loss": 1.418,
+      "step": 18799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.65298594395938,
+      "learning_rate": 1.5530210115083403e-06,
+      "loss": 1.0996,
+      "step": 18800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.146931002942304,
+      "learning_rate": 1.5515743147441586e-06,
+      "loss": 0.8008,
+      "step": 18801
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7626936728134495,
+      "learning_rate": 1.5501282705401104e-06,
+      "loss": 1.1445,
+      "step": 18802
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.508173075865642,
+      "learning_rate": 1.54868287893645e-06,
+      "loss": 1.2891,
+      "step": 18803
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.737682634924756,
+      "learning_rate": 1.5472381399734005e-06,
+      "loss": 1.457,
+      "step": 18804
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7537511501761447,
+      "learning_rate": 1.5457940536911658e-06,
+      "loss": 1.1133,
+      "step": 18805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.865302000797069,
+      "learning_rate": 1.5443506201299279e-06,
+      "loss": 1.498,
+      "step": 18806
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.1109738806527005,
+      "learning_rate": 1.5429078393298658e-06,
+      "loss": 1.1367,
+      "step": 18807
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9812789553863956,
+      "learning_rate": 1.5414657113311304e-06,
+      "loss": 0.7617,
+      "step": 18808
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.099631177212385,
+      "learning_rate": 1.5400242361738566e-06,
+      "loss": 0.5918,
+      "step": 18809
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.300277473463916,
+      "learning_rate": 1.5385834138981536e-06,
+      "loss": 0.6851,
+      "step": 18810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.318878275156913,
+      "learning_rate": 1.5371432445441257e-06,
+      "loss": 1.4102,
+      "step": 18811
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.445642451308281,
+      "learning_rate": 1.5357037281518522e-06,
+      "loss": 0.7822,
+      "step": 18812
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.258950745867335,
+      "learning_rate": 1.5342648647613893e-06,
+      "loss": 1.0708,
+      "step": 18813
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.470993567723509,
+      "learning_rate": 1.5328266544127779e-06,
+      "loss": 1.5918,
+      "step": 18814
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1299336325465013,
+      "learning_rate": 1.531389097146052e-06,
+      "loss": 0.9844,
+      "step": 18815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.34054101607402,
+      "learning_rate": 1.5299521930012107e-06,
+      "loss": 0.4414,
+      "step": 18816
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.659413721415613,
+      "learning_rate": 1.5285159420182494e-06,
+      "loss": 1.4453,
+      "step": 18817
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.875506149187247,
+      "learning_rate": 1.5270803442371307e-06,
+      "loss": 1.124,
+      "step": 18818
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5130696082776107,
+      "learning_rate": 1.5256453996978087e-06,
+      "loss": 0.6479,
+      "step": 18819
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.810035534390123,
+      "learning_rate": 1.5242111084402238e-06,
+      "loss": 0.9331,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3418339738954765,
+      "learning_rate": 1.5227774705042858e-06,
+      "loss": 0.9314,
+      "step": 18821
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.2213938023069195,
+      "learning_rate": 1.5213444859298932e-06,
+      "loss": 1.0098,
+      "step": 18822
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.331598643793431,
+      "learning_rate": 1.5199121547569195e-06,
+      "loss": 0.8428,
+      "step": 18823
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1579927782109345,
+      "learning_rate": 1.5184804770252386e-06,
+      "loss": 1.2266,
+      "step": 18824
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.962922611894226,
+      "learning_rate": 1.5170494527746826e-06,
+      "loss": 1.5039,
+      "step": 18825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9373725110840458,
+      "learning_rate": 1.5156190820450832e-06,
+      "loss": 0.8818,
+      "step": 18826
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.4319681422331465,
+      "learning_rate": 1.5141893648762395e-06,
+      "loss": 1.0781,
+      "step": 18827
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3467273366013988,
+      "learning_rate": 1.51276030130795e-06,
+      "loss": 1.1035,
+      "step": 18828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0929403305260124,
+      "learning_rate": 1.51133189137998e-06,
+      "loss": 1.2295,
+      "step": 18829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.018264816133401,
+      "learning_rate": 1.5099041351320787e-06,
+      "loss": 1.0283,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.614959572080083,
+      "learning_rate": 1.5084770326039804e-06,
+      "loss": 1.1426,
+      "step": 18831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.366683461994038,
+      "learning_rate": 1.5070505838354065e-06,
+      "loss": 1.123,
+      "step": 18832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.23863333956075,
+      "learning_rate": 1.505624788866053e-06,
+      "loss": 1.1191,
+      "step": 18833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.737959609471816,
+      "learning_rate": 1.5041996477355935e-06,
+      "loss": 1.0205,
+      "step": 18834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.397908503443668,
+      "learning_rate": 1.5027751604836909e-06,
+      "loss": 1.2725,
+      "step": 18835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.302194468949956,
+      "learning_rate": 1.5013513271499884e-06,
+      "loss": 1.0215,
+      "step": 18836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.064873786400602,
+      "learning_rate": 1.4999281477741184e-06,
+      "loss": 1.3203,
+      "step": 18837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.26431794446619,
+      "learning_rate": 1.4985056223956796e-06,
+      "loss": 0.9775,
+      "step": 18838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4352908760872003,
+      "learning_rate": 1.4970837510542623e-06,
+      "loss": 0.8301,
+      "step": 18839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9212920071377315,
+      "learning_rate": 1.4956625337894324e-06,
+      "loss": 1.666,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3412589195944085,
+      "learning_rate": 1.4942419706407474e-06,
+      "loss": 1.0615,
+      "step": 18841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.163343620021131,
+      "learning_rate": 1.4928220616477418e-06,
+      "loss": 1.0098,
+      "step": 18842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.691497225393431,
+      "learning_rate": 1.4914028068499286e-06,
+      "loss": 1.4238,
+      "step": 18843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4587168416338252,
+      "learning_rate": 1.4899842062867958e-06,
+      "loss": 0.1808,
+      "step": 18844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.706774854972063,
+      "learning_rate": 1.4885662599978395e-06,
+      "loss": 1.3594,
+      "step": 18845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.348758581755554,
+      "learning_rate": 1.4871489680225115e-06,
+      "loss": 0.9177,
+      "step": 18846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.01331984778532,
+      "learning_rate": 1.485732330400258e-06,
+      "loss": 1.5664,
+      "step": 18847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7275712989010836,
+      "learning_rate": 1.4843163471704974e-06,
+      "loss": 1.29,
+      "step": 18848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.526489270919736,
+      "learning_rate": 1.48290101837264e-06,
+      "loss": 1.6973,
+      "step": 18849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.153090803638491,
+      "learning_rate": 1.4814863440460764e-06,
+      "loss": 0.9014,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.790035791353909,
+      "learning_rate": 1.4800723242301723e-06,
+      "loss": 0.8003,
+      "step": 18851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3486259095532653,
+      "learning_rate": 1.4786589589642768e-06,
+      "loss": 0.4749,
+      "step": 18852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.81458485018834,
+      "learning_rate": 1.477246248287728e-06,
+      "loss": 1.6992,
+      "step": 18853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.092556810034541,
+      "learning_rate": 1.4758341922398444e-06,
+      "loss": 1.4004,
+      "step": 18854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8224082310391057,
+      "learning_rate": 1.4744227908599168e-06,
+      "loss": 0.8467,
+      "step": 18855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0979667447984665,
+      "learning_rate": 1.4730120441872248e-06,
+      "loss": 1.627,
+      "step": 18856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.348550718228832,
+      "learning_rate": 1.4716019522610286e-06,
+      "loss": 1.5938,
+      "step": 18857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0645922827759238,
+      "learning_rate": 1.470192515120572e-06,
+      "loss": 0.4835,
+      "step": 18858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.720647040430551,
+      "learning_rate": 1.4687837328050818e-06,
+      "loss": 0.751,
+      "step": 18859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0998963057938527,
+      "learning_rate": 1.4673756053537602e-06,
+      "loss": 1.3945,
+      "step": 18860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5462398548539125,
+      "learning_rate": 1.4659681328057895e-06,
+      "loss": 1.0703,
+      "step": 18861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.146136240890902,
+      "learning_rate": 1.4645613152003469e-06,
+      "loss": 1.1104,
+      "step": 18862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.925982296490932,
+      "learning_rate": 1.463155152576587e-06,
+      "loss": 0.7825,
+      "step": 18863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.770453768859568,
+      "learning_rate": 1.4617496449736339e-06,
+      "loss": 1.4688,
+      "step": 18864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8047586501083592,
+      "learning_rate": 1.4603447924306041e-06,
+      "loss": 0.6396,
+      "step": 18865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9811601804957266,
+      "learning_rate": 1.4589405949865992e-06,
+      "loss": 1.3965,
+      "step": 18866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.799234854513182,
+      "learning_rate": 1.4575370526806936e-06,
+      "loss": 0.7002,
+      "step": 18867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.732136841754146,
+      "learning_rate": 1.456134165551948e-06,
+      "loss": 1.1182,
+      "step": 18868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6499426798042918,
+      "learning_rate": 1.4547319336394e-06,
+      "loss": 0.75,
+      "step": 18869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.695401397565907,
+      "learning_rate": 1.45333035698208e-06,
+      "loss": 0.4795,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.642258766145113,
+      "learning_rate": 1.45192943561899e-06,
+      "loss": 0.3252,
+      "step": 18871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8943592111400887,
+      "learning_rate": 1.4505291695891154e-06,
+      "loss": 1.1475,
+      "step": 18872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3413310474516384,
+      "learning_rate": 1.4491295589314308e-06,
+      "loss": 0.9346,
+      "step": 18873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.629273514665225,
+      "learning_rate": 1.447730603684877e-06,
+      "loss": 1.2363,
+      "step": 18874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.192446392082823,
+      "learning_rate": 1.446332303888398e-06,
+      "loss": 1.0166,
+      "step": 18875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.578379300609214,
+      "learning_rate": 1.4449346595809015e-06,
+      "loss": 1.29,
+      "step": 18876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.607855077464675,
+      "learning_rate": 1.4435376708012844e-06,
+      "loss": 0.6016,
+      "step": 18877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3296567243235717,
+      "learning_rate": 1.4421413375884207e-06,
+      "loss": 1.6465,
+      "step": 18878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.403492920965313,
+      "learning_rate": 1.4407456599811769e-06,
+      "loss": 1.1895,
+      "step": 18879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.96152570946908,
+      "learning_rate": 1.4393506380183885e-06,
+      "loss": 1.3633,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.309641899933689,
+      "learning_rate": 1.43795627173888e-06,
+      "loss": 0.3887,
+      "step": 18881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5498519919086209,
+      "learning_rate": 1.4365625611814564e-06,
+      "loss": 0.5753,
+      "step": 18882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9436603372722057,
+      "learning_rate": 1.4351695063849036e-06,
+      "loss": 1.3105,
+      "step": 18883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.302898889671673,
+      "learning_rate": 1.4337771073879902e-06,
+      "loss": 0.8848,
+      "step": 18884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.204001534270725,
+      "learning_rate": 1.4323853642294687e-06,
+      "loss": 1.8398,
+      "step": 18885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.67677300435385,
+      "learning_rate": 1.4309942769480638e-06,
+      "loss": 1.4004,
+      "step": 18886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.454096316613206,
+      "learning_rate": 1.4296038455824973e-06,
+      "loss": 1.1587,
+      "step": 18887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.258034137552247,
+      "learning_rate": 1.4282140701714602e-06,
+      "loss": 0.8284,
+      "step": 18888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.785678136803459,
+      "learning_rate": 1.4268249507536301e-06,
+      "loss": 1.1035,
+      "step": 18889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.493568474696719,
+      "learning_rate": 1.4254364873676623e-06,
+      "loss": 1.0059,
+      "step": 18890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2519177867355458,
+      "learning_rate": 1.424048680052198e-06,
+      "loss": 0.9473,
+      "step": 18891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.497033075286979,
+      "learning_rate": 1.4226615288458644e-06,
+      "loss": 0.79,
+      "step": 18892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.11637832142126,
+      "learning_rate": 1.4212750337872617e-06,
+      "loss": 1.3223,
+      "step": 18893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.162423625325012,
+      "learning_rate": 1.4198891949149778e-06,
+      "loss": 1.291,
+      "step": 18894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.159672480929535,
+      "learning_rate": 1.4185040122675741e-06,
+      "loss": 1.5801,
+      "step": 18895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8140742601379818,
+      "learning_rate": 1.4171194858836057e-06,
+      "loss": 0.5964,
+      "step": 18896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.034161130169434,
+      "learning_rate": 1.4157356158016027e-06,
+      "loss": 0.9814,
+      "step": 18897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.812583159170978,
+      "learning_rate": 1.4143524020600763e-06,
+      "loss": 1.2012,
+      "step": 18898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.416028848425312,
+      "learning_rate": 1.4129698446975176e-06,
+      "loss": 1.0078,
+      "step": 18899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8037111075735757,
+      "learning_rate": 1.4115879437524043e-06,
+      "loss": 1.0676,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4806157106132853,
+      "learning_rate": 1.4102066992632002e-06,
+      "loss": 1.0615,
+      "step": 18901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.961448283161111,
+      "learning_rate": 1.408826111268341e-06,
+      "loss": 0.5405,
+      "step": 18902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.971932155548569,
+      "learning_rate": 1.407446179806246e-06,
+      "loss": 0.9395,
+      "step": 18903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.616767518678606,
+      "learning_rate": 1.4060669049153152e-06,
+      "loss": 1.4424,
+      "step": 18904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9813784705065414,
+      "learning_rate": 1.404688286633943e-06,
+      "loss": 1.2168,
+      "step": 18905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.213393825006941,
+      "learning_rate": 1.4033103250004898e-06,
+      "loss": 0.708,
+      "step": 18906
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2925858705967936,
+      "learning_rate": 1.4019330200533004e-06,
+      "loss": 0.9424,
+      "step": 18907
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9492769551375586,
+      "learning_rate": 1.4005563718307108e-06,
+      "loss": 1.2051,
+      "step": 18908
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3823295407083327,
+      "learning_rate": 1.3991803803710263e-06,
+      "loss": 0.7485,
+      "step": 18909
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.336255525174931,
+      "learning_rate": 1.397805045712547e-06,
+      "loss": 0.6714,
+      "step": 18910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4015059726207895,
+      "learning_rate": 1.3964303678935476e-06,
+      "loss": 1.1199,
+      "step": 18911
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.096489324220914,
+      "learning_rate": 1.3950563469522753e-06,
+      "loss": 0.6111,
+      "step": 18912
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0104609140326573,
+      "learning_rate": 1.3936829829269804e-06,
+      "loss": 0.7856,
+      "step": 18913
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.30084283065393,
+      "learning_rate": 1.3923102758558792e-06,
+      "loss": 0.959,
+      "step": 18914
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3939692225059335,
+      "learning_rate": 1.3909382257771692e-06,
+      "loss": 1.0293,
+      "step": 18915
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.876907213599636,
+      "learning_rate": 1.3895668327290361e-06,
+      "loss": 0.7612,
+      "step": 18916
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.96615773226173,
+      "learning_rate": 1.38819609674965e-06,
+      "loss": 1.9863,
+      "step": 18917
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7076740931261887,
+      "learning_rate": 1.3868260178771492e-06,
+      "loss": 0.755,
+      "step": 18918
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.404934351186444,
+      "learning_rate": 1.38545659614967e-06,
+      "loss": 1.3203,
+      "step": 18919
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.241897298395159,
+      "learning_rate": 1.384087831605324e-06,
+      "loss": 0.749,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.316127633471129,
+      "learning_rate": 1.3827197242821914e-06,
+      "loss": 1.1719,
+      "step": 18921
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3748546031224755,
+      "learning_rate": 1.3813522742183616e-06,
+      "loss": 1.0098,
+      "step": 18922
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0893021794958475,
+      "learning_rate": 1.3799854814518814e-06,
+      "loss": 0.9121,
+      "step": 18923
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9803427631701798,
+      "learning_rate": 1.3786193460207846e-06,
+      "loss": 0.8311,
+      "step": 18924
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.119577318568031,
+      "learning_rate": 1.377253867963102e-06,
+      "loss": 0.5864,
+      "step": 18925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3025140373804422,
+      "learning_rate": 1.375889047316825e-06,
+      "loss": 0.8818,
+      "step": 18926
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1989003250756776,
+      "learning_rate": 1.3745248841199348e-06,
+      "loss": 0.9795,
+      "step": 18927
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4406594560953074,
+      "learning_rate": 1.3731613784104037e-06,
+      "loss": 1.0342,
+      "step": 18928
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7338830061034598,
+      "learning_rate": 1.3717985302261676e-06,
+      "loss": 1.0205,
+      "step": 18929
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.795461440173859,
+      "learning_rate": 1.3704363396051661e-06,
+      "loss": 0.7096,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8571104660922897,
+      "learning_rate": 1.3690748065852994e-06,
+      "loss": 0.8591,
+      "step": 18931
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2998311317470286,
+      "learning_rate": 1.3677139312044618e-06,
+      "loss": 0.7559,
+      "step": 18932
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.540839503618478,
+      "learning_rate": 1.3663537135005178e-06,
+      "loss": 0.8311,
+      "step": 18933
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9066196771091104,
+      "learning_rate": 1.364994153511337e-06,
+      "loss": 0.3826,
+      "step": 18934
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.090732392658999,
+      "learning_rate": 1.3636352512747447e-06,
+      "loss": 0.731,
+      "step": 18935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9549240371697634,
+      "learning_rate": 1.3622770068285607e-06,
+      "loss": 0.8706,
+      "step": 18936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.597310391684435,
+      "learning_rate": 1.3609194202105824e-06,
+      "loss": 0.7144,
+      "step": 18937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.059769795304272,
+      "learning_rate": 1.3595624914585937e-06,
+      "loss": 1.3945,
+      "step": 18938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.612901179667246,
+      "learning_rate": 1.3582062206103585e-06,
+      "loss": 1.1333,
+      "step": 18939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.619945547930703,
+      "learning_rate": 1.3568506077036192e-06,
+      "loss": 1.0767,
+      "step": 18940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.869905283356891,
+      "learning_rate": 1.3554956527761038e-06,
+      "loss": 1.2676,
+      "step": 18941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9410087245901253,
+      "learning_rate": 1.3541413558655153e-06,
+      "loss": 0.5269,
+      "step": 18942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.379085951713279,
+      "learning_rate": 1.3527877170095488e-06,
+      "loss": 1.2891,
+      "step": 18943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.503161615124727,
+      "learning_rate": 1.351434736245874e-06,
+      "loss": 1.1553,
+      "step": 18944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.551799967958519,
+      "learning_rate": 1.3500824136121414e-06,
+      "loss": 1.0449,
+      "step": 18945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.955821748647545,
+      "learning_rate": 1.348730749145985e-06,
+      "loss": 0.8252,
+      "step": 18946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.2777587905341585,
+      "learning_rate": 1.347379742885027e-06,
+      "loss": 1.2275,
+      "step": 18947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.089529876178226,
+      "learning_rate": 1.3460293948668629e-06,
+      "loss": 0.6914,
+      "step": 18948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7500086697528463,
+      "learning_rate": 1.3446797051290704e-06,
+      "loss": 0.6396,
+      "step": 18949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.977836888866799,
+      "learning_rate": 1.343330673709209e-06,
+      "loss": 1.0029,
+      "step": 18950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.608394232903294,
+      "learning_rate": 1.3419823006448284e-06,
+      "loss": 0.8643,
+      "step": 18951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.854484503282507,
+      "learning_rate": 1.3406345859734493e-06,
+      "loss": 0.9209,
+      "step": 18952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4566849068732717,
+      "learning_rate": 1.3392875297325747e-06,
+      "loss": 1.1455,
+      "step": 18953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9951811195321463,
+      "learning_rate": 1.337941131959694e-06,
+      "loss": 0.9565,
+      "step": 18954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3060298565152753,
+      "learning_rate": 1.3365953926922803e-06,
+      "loss": 0.6777,
+      "step": 18955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.293801995288648,
+      "learning_rate": 1.3352503119677866e-06,
+      "loss": 1.1172,
+      "step": 18956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.473959233151862,
+      "learning_rate": 1.3339058898236416e-06,
+      "loss": 1.2051,
+      "step": 18957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.458608801119954,
+      "learning_rate": 1.3325621262972626e-06,
+      "loss": 0.8537,
+      "step": 18958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.782235237286649,
+      "learning_rate": 1.3312190214260389e-06,
+      "loss": 1.4824,
+      "step": 18959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.623882267421012,
+      "learning_rate": 1.32987657524736e-06,
+      "loss": 0.9248,
+      "step": 18960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8065796206710694,
+      "learning_rate": 1.3285347877985799e-06,
+      "loss": 1.1187,
+      "step": 18961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.62695847439748,
+      "learning_rate": 1.3271936591170375e-06,
+      "loss": 1.7129,
+      "step": 18962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0537066421064805,
+      "learning_rate": 1.3258531892400561e-06,
+      "loss": 1.3301,
+      "step": 18963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3666277993353844,
+      "learning_rate": 1.3245133782049445e-06,
+      "loss": 0.8049,
+      "step": 18964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4205418024359933,
+      "learning_rate": 1.323174226048987e-06,
+      "loss": 1.3789,
+      "step": 18965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.86701835944182,
+      "learning_rate": 1.3218357328094567e-06,
+      "loss": 1.6016,
+      "step": 18966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4856206302350228,
+      "learning_rate": 1.3204978985235927e-06,
+      "loss": 1.2695,
+      "step": 18967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0650963190324525,
+      "learning_rate": 1.319160723228635e-06,
+      "loss": 1.0449,
+      "step": 18968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.843365409653256,
+      "learning_rate": 1.317824206961793e-06,
+      "loss": 0.7207,
+      "step": 18969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.483717813751174,
+      "learning_rate": 1.3164883497602642e-06,
+      "loss": 0.7041,
+      "step": 18970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.931949159574329,
+      "learning_rate": 1.3151531516612192e-06,
+      "loss": 0.9121,
+      "step": 18971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7197551215864009,
+      "learning_rate": 1.3138186127018254e-06,
+      "loss": 0.8415,
+      "step": 18972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1517953606769757,
+      "learning_rate": 1.3124847329192142e-06,
+      "loss": 1.0417,
+      "step": 18973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3671407674214984,
+      "learning_rate": 1.3111515123505057e-06,
+      "loss": 1.1924,
+      "step": 18974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.700466242267737,
+      "learning_rate": 1.3098189510328123e-06,
+      "loss": 1.0371,
+      "step": 18975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.1808543881908085,
+      "learning_rate": 1.3084870490032098e-06,
+      "loss": 1.5801,
+      "step": 18976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2994730326357904,
+      "learning_rate": 1.307155806298771e-06,
+      "loss": 1.5312,
+      "step": 18977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.087140199028783,
+      "learning_rate": 1.3058252229565415e-06,
+      "loss": 1.0596,
+      "step": 18978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.51766298351627,
+      "learning_rate": 1.3044952990135472e-06,
+      "loss": 0.9424,
+      "step": 18979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9191398696696051,
+      "learning_rate": 1.3031660345068002e-06,
+      "loss": 1.0645,
+      "step": 18980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.272387704624137,
+      "learning_rate": 1.3018374294733016e-06,
+      "loss": 0.6489,
+      "step": 18981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.66291342531024,
+      "learning_rate": 1.300509483950016e-06,
+      "loss": 1.1504,
+      "step": 18982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1783415764689416,
+      "learning_rate": 1.2991821979739027e-06,
+      "loss": 1.0879,
+      "step": 18983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.178006416059139,
+      "learning_rate": 1.2978555715818991e-06,
+      "loss": 1.252,
+      "step": 18984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.7505602356363195,
+      "learning_rate": 1.2965296048109282e-06,
+      "loss": 0.8091,
+      "step": 18985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.581428169225685,
+      "learning_rate": 1.2952042976978911e-06,
+      "loss": 0.7451,
+      "step": 18986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1664065873272706,
+      "learning_rate": 1.2938796502796669e-06,
+      "loss": 1.293,
+      "step": 18987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8978102185088126,
+      "learning_rate": 1.2925556625931173e-06,
+      "loss": 0.927,
+      "step": 18988
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.445874816785349,
+      "learning_rate": 1.2912323346750965e-06,
+      "loss": 1.0684,
+      "step": 18989
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9926218195186833,
+      "learning_rate": 1.2899096665624278e-06,
+      "loss": 0.6082,
+      "step": 18990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.2913012055770805,
+      "learning_rate": 1.2885876582919203e-06,
+      "loss": 1.2871,
+      "step": 18991
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2022429519975315,
+      "learning_rate": 1.2872663099003617e-06,
+      "loss": 1.4961,
+      "step": 18992
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.346677553935516,
+      "learning_rate": 1.285945621424528e-06,
+      "loss": 0.8867,
+      "step": 18993
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5991037107692945,
+      "learning_rate": 1.2846255929011785e-06,
+      "loss": 1.041,
+      "step": 18994
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.259259627214744,
+      "learning_rate": 1.2833062243670452e-06,
+      "loss": 0.6049,
+      "step": 18995
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5740885392948436,
+      "learning_rate": 1.281987515858843e-06,
+      "loss": 1.2891,
+      "step": 18996
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9463977773787746,
+      "learning_rate": 1.2806694674132673e-06,
+      "loss": 0.728,
+      "step": 18997
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.501685718596832,
+      "learning_rate": 1.2793520790670116e-06,
+      "loss": 0.9375,
+      "step": 18998
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1454984780179283,
+      "learning_rate": 1.2780353508567266e-06,
+      "loss": 0.8583,
+      "step": 18999
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6447724045037893,
+      "learning_rate": 1.2767192828190638e-06,
+      "loss": 1.125,
+      "step": 19000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.02300406524597,
+      "learning_rate": 1.2754038749906382e-06,
+      "loss": 1.5469,
+      "step": 19001
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9505852975337725,
+      "learning_rate": 1.2740891274080707e-06,
+      "loss": 0.4463,
+      "step": 19002
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4386023938486616,
+      "learning_rate": 1.272775040107943e-06,
+      "loss": 1.1318,
+      "step": 19003
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.506100744811699,
+      "learning_rate": 1.2714616131268286e-06,
+      "loss": 1.623,
+      "step": 19004
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5038002693968986,
+      "learning_rate": 1.2701488465012733e-06,
+      "loss": 1.293,
+      "step": 19005
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.929907487357927,
+      "learning_rate": 1.2688367402678175e-06,
+      "loss": 1.0176,
+      "step": 19006
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0231239341000027,
+      "learning_rate": 1.2675252944629761e-06,
+      "loss": 1.1523,
+      "step": 19007
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.798092625004442,
+      "learning_rate": 1.2662145091232425e-06,
+      "loss": 1.1353,
+      "step": 19008
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.376506227407677,
+      "learning_rate": 1.2649043842850956e-06,
+      "loss": 1.0234,
+      "step": 19009
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6557491054610614,
+      "learning_rate": 1.2635949199849978e-06,
+      "loss": 1.3193,
+      "step": 19010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.221493190930477,
+      "learning_rate": 1.2622861162593897e-06,
+      "loss": 1.377,
+      "step": 19011
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.013945116572372,
+      "learning_rate": 1.2609779731446975e-06,
+      "loss": 0.9678,
+      "step": 19012
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.431371620994946,
+      "learning_rate": 1.2596704906773255e-06,
+      "loss": 1.0098,
+      "step": 19013
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.494902468286772,
+      "learning_rate": 1.2583636688936528e-06,
+      "loss": 1.1201,
+      "step": 19014
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4818745631782186,
+      "learning_rate": 1.257057507830059e-06,
+      "loss": 1.1567,
+      "step": 19015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7283335062253453,
+      "learning_rate": 1.2557520075228895e-06,
+      "loss": 0.9722,
+      "step": 19016
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9615799660700437,
+      "learning_rate": 1.254447168008477e-06,
+      "loss": 0.7451,
+      "step": 19017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.998597666724874,
+      "learning_rate": 1.253142989323125e-06,
+      "loss": 0.9277,
+      "step": 19018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5810112720407417,
+      "learning_rate": 1.2518394715031439e-06,
+      "loss": 0.9141,
+      "step": 19019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2729455999064907,
+      "learning_rate": 1.2505366145847964e-06,
+      "loss": 0.9609,
+      "step": 19020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6547777471531933,
+      "learning_rate": 1.2492344186043503e-06,
+      "loss": 0.7676,
+      "step": 19021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3476407241016135,
+      "learning_rate": 1.247932883598038e-06,
+      "loss": 1.3076,
+      "step": 19022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.723990756034408,
+      "learning_rate": 1.2466320096020888e-06,
+      "loss": 0.705,
+      "step": 19023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8247888604315308,
+      "learning_rate": 1.245331796652699e-06,
+      "loss": 0.9917,
+      "step": 19024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.60348293229544,
+      "learning_rate": 1.2440322447860558e-06,
+      "loss": 1.541,
+      "step": 19025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3386438314866456,
+      "learning_rate": 1.2427333540383224e-06,
+      "loss": 0.8174,
+      "step": 19026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.324846638263934,
+      "learning_rate": 1.24143512444565e-06,
+      "loss": 1.0547,
+      "step": 19027
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.9445094444706195,
+      "learning_rate": 1.2401375560441658e-06,
+      "loss": 1.5645,
+      "step": 19028
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.010087882432142,
+      "learning_rate": 1.2388406488699767e-06,
+      "loss": 1.3438,
+      "step": 19029
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.191490348521379,
+      "learning_rate": 1.2375444029591843e-06,
+      "loss": 0.7679,
+      "step": 19030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6932119836789643,
+      "learning_rate": 1.2362488183478516e-06,
+      "loss": 1.1738,
+      "step": 19031
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.657309198097817,
+      "learning_rate": 1.2349538950720441e-06,
+      "loss": 2.0566,
+      "step": 19032
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2973179811720037,
+      "learning_rate": 1.233659633167797e-06,
+      "loss": 0.8633,
+      "step": 19033
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.140804911911904,
+      "learning_rate": 1.2323660326711257e-06,
+      "loss": 1.2656,
+      "step": 19034
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2146745083319157,
+      "learning_rate": 1.2310730936180292e-06,
+      "loss": 1.5303,
+      "step": 19035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0566891975137316,
+      "learning_rate": 1.2297808160444929e-06,
+      "loss": 0.7676,
+      "step": 19036
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.180345030785858,
+      "learning_rate": 1.2284891999864822e-06,
+      "loss": 1.0547,
+      "step": 19037
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.606140141721749,
+      "learning_rate": 1.2271982454799407e-06,
+      "loss": 0.7539,
+      "step": 19038
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.200561986689158,
+      "learning_rate": 1.2259079525607869e-06,
+      "loss": 1.4902,
+      "step": 19039
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3346307772606996,
+      "learning_rate": 1.2246183212649392e-06,
+      "loss": 1.3242,
+      "step": 19040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8633326237471883,
+      "learning_rate": 1.2233293516282884e-06,
+      "loss": 0.792,
+      "step": 19041
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5253876039231735,
+      "learning_rate": 1.2220410436867002e-06,
+      "loss": 0.9688,
+      "step": 19042
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7759503223690336,
+      "learning_rate": 1.2207533974760294e-06,
+      "loss": 1.0142,
+      "step": 19043
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0038386699740083,
+      "learning_rate": 1.219466413032111e-06,
+      "loss": 0.9568,
+      "step": 19044
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8282430887799777,
+      "learning_rate": 1.2181800903907636e-06,
+      "loss": 0.2476,
+      "step": 19045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.685616280032548,
+      "learning_rate": 1.216894429587781e-06,
+      "loss": 1.3945,
+      "step": 19046
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4499367569526416,
+      "learning_rate": 1.2156094306589427e-06,
+      "loss": 1.2695,
+      "step": 19047
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4071920953414816,
+      "learning_rate": 1.214325093640009e-06,
+      "loss": 1.1055,
+      "step": 19048
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6556411627498417,
+      "learning_rate": 1.2130414185667294e-06,
+      "loss": 1.2266,
+      "step": 19049
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3041011775702716,
+      "learning_rate": 1.2117584054748248e-06,
+      "loss": 1.207,
+      "step": 19050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.926387401678932,
+      "learning_rate": 1.2104760543999976e-06,
+      "loss": 0.71,
+      "step": 19051
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0214228263163934,
+      "learning_rate": 1.209194365377936e-06,
+      "loss": 1.3418,
+      "step": 19052
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7917706674210647,
+      "learning_rate": 1.207913338444311e-06,
+      "loss": 0.792,
+      "step": 19053
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.875209966538891,
+      "learning_rate": 1.206632973634772e-06,
+      "loss": 1.5117,
+      "step": 19054
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4890067631287454,
+      "learning_rate": 1.205353270984952e-06,
+      "loss": 1.151,
+      "step": 19055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.733206624684009,
+      "learning_rate": 1.204074230530461e-06,
+      "loss": 1.0612,
+      "step": 19056
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.933195653487538,
+      "learning_rate": 1.2027958523068954e-06,
+      "loss": 1.0049,
+      "step": 19057
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3841971954805365,
+      "learning_rate": 1.2015181363498384e-06,
+      "loss": 1.1367,
+      "step": 19058
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.234134831977207,
+      "learning_rate": 1.2002410826948418e-06,
+      "loss": 1.5215,
+      "step": 19059
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.315384913755196,
+      "learning_rate": 1.1989646913774466e-06,
+      "loss": 1.4844,
+      "step": 19060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1829520298407856,
+      "learning_rate": 1.1976889624331743e-06,
+      "loss": 0.8896,
+      "step": 19061
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.1699456600609714,
+      "learning_rate": 1.1964138958975273e-06,
+      "loss": 0.9033,
+      "step": 19062
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.315685346591839,
+      "learning_rate": 1.1951394918059938e-06,
+      "loss": 0.7871,
+      "step": 19063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.393041848299217,
+      "learning_rate": 1.1938657501940343e-06,
+      "loss": 0.6194,
+      "step": 19064
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.046486854151247,
+      "learning_rate": 1.192592671097101e-06,
+      "loss": 1.0742,
+      "step": 19065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0157064318162585,
+      "learning_rate": 1.1913202545506209e-06,
+      "loss": 1.0547,
+      "step": 19066
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1279652069771444,
+      "learning_rate": 1.1900485005900076e-06,
+      "loss": 0.5586,
+      "step": 19067
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.061860958602779,
+      "learning_rate": 1.1887774092506494e-06,
+      "loss": 1.166,
+      "step": 19068
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2103917826678923,
+      "learning_rate": 1.1875069805679207e-06,
+      "loss": 1.1875,
+      "step": 19069
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.639576186523049,
+      "learning_rate": 1.186237214577182e-06,
+      "loss": 0.9492,
+      "step": 19070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.045658905997428,
+      "learning_rate": 1.1849681113137668e-06,
+      "loss": 1.2595,
+      "step": 19071
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0390931342735916,
+      "learning_rate": 1.183699670812996e-06,
+      "loss": 0.9336,
+      "step": 19072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1080955863866264,
+      "learning_rate": 1.1824318931101614e-06,
+      "loss": 1.1328,
+      "step": 19073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7831417427341623,
+      "learning_rate": 1.1811647782405567e-06,
+      "loss": 0.9473,
+      "step": 19074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.946193573402883,
+      "learning_rate": 1.1798983262394397e-06,
+      "loss": 0.9033,
+      "step": 19075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5361730479095606,
+      "learning_rate": 1.1786325371420493e-06,
+      "loss": 1.083,
+      "step": 19076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6152736669413446,
+      "learning_rate": 1.1773674109836237e-06,
+      "loss": 0.6567,
+      "step": 19077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.759010275798593,
+      "learning_rate": 1.1761029477993597e-06,
+      "loss": 0.8057,
+      "step": 19078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.83629362547013,
+      "learning_rate": 1.174839147624454e-06,
+      "loss": 0.9219,
+      "step": 19079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.916668337866895,
+      "learning_rate": 1.173576010494079e-06,
+      "loss": 0.4575,
+      "step": 19080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.525841763376507,
+      "learning_rate": 1.1723135364433785e-06,
+      "loss": 1.4941,
+      "step": 19081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7055927103908792,
+      "learning_rate": 1.1710517255074966e-06,
+      "loss": 0.6123,
+      "step": 19082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4248315300379453,
+      "learning_rate": 1.169790577721544e-06,
+      "loss": 1.2266,
+      "step": 19083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.629942373797542,
+      "learning_rate": 1.1685300931206178e-06,
+      "loss": 1.0957,
+      "step": 19084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6064924205126125,
+      "learning_rate": 1.1672702717397927e-06,
+      "loss": 0.8892,
+      "step": 19085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7423939180087427,
+      "learning_rate": 1.166011113614135e-06,
+      "loss": 1.1104,
+      "step": 19086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.510953653185101,
+      "learning_rate": 1.164752618778689e-06,
+      "loss": 0.8418,
+      "step": 19087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.623737718779896,
+      "learning_rate": 1.1634947872684737e-06,
+      "loss": 1.7109,
+      "step": 19088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0647818392574075,
+      "learning_rate": 1.1622376191184947e-06,
+      "loss": 1.0371,
+      "step": 19089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0424062970541885,
+      "learning_rate": 1.160981114363735e-06,
+      "loss": 1.0361,
+      "step": 19090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.983736459216303,
+      "learning_rate": 1.1597252730391695e-06,
+      "loss": 0.8838,
+      "step": 19091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.826730879813009,
+      "learning_rate": 1.158470095179745e-06,
+      "loss": 0.8926,
+      "step": 19092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8123722993069484,
+      "learning_rate": 1.1572155808203922e-06,
+      "loss": 1.0186,
+      "step": 19093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4478692150404195,
+      "learning_rate": 1.155961729996019e-06,
+      "loss": 1.1309,
+      "step": 19094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.392776518400851,
+      "learning_rate": 1.1547085427415227e-06,
+      "loss": 1.1211,
+      "step": 19095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.806735765175906,
+      "learning_rate": 1.1534560190917865e-06,
+      "loss": 0.5293,
+      "step": 19096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4718265919339975,
+      "learning_rate": 1.1522041590816602e-06,
+      "loss": 1.6504,
+      "step": 19097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5038982516204156,
+      "learning_rate": 1.1509529627459825e-06,
+      "loss": 0.4521,
+      "step": 19098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9305109282655801,
+      "learning_rate": 1.149702430119573e-06,
+      "loss": 1.3008,
+      "step": 19099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.691528937612073,
+      "learning_rate": 1.148452561237237e-06,
+      "loss": 1.3086,
+      "step": 19100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5893370994392506,
+      "learning_rate": 1.147203356133758e-06,
+      "loss": 0.894,
+      "step": 19101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.101942876707272,
+      "learning_rate": 1.1459548148438942e-06,
+      "loss": 0.7412,
+      "step": 19102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6519033434358095,
+      "learning_rate": 1.1447069374023983e-06,
+      "loss": 1.0547,
+      "step": 19103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.845808839463435,
+      "learning_rate": 1.143459723843998e-06,
+      "loss": 1.0059,
+      "step": 19104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7015338811255147,
+      "learning_rate": 1.1422131742034048e-06,
+      "loss": 1.2715,
+      "step": 19105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2795303925008708,
+      "learning_rate": 1.1409672885153044e-06,
+      "loss": 0.7454,
+      "step": 19106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.639707542321318,
+      "learning_rate": 1.1397220668143693e-06,
+      "loss": 1.231,
+      "step": 19107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5393731029971915,
+      "learning_rate": 1.1384775091352579e-06,
+      "loss": 1.001,
+      "step": 19108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.891441862329635,
+      "learning_rate": 1.1372336155126033e-06,
+      "loss": 0.8337,
+      "step": 19109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5261975951424245,
+      "learning_rate": 1.1359903859810227e-06,
+      "loss": 0.8672,
+      "step": 19110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.468111358240293,
+      "learning_rate": 1.1347478205751105e-06,
+      "loss": 1.2231,
+      "step": 19111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.061935372755067,
+      "learning_rate": 1.1335059193294556e-06,
+      "loss": 0.7998,
+      "step": 19112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1848011261544946,
+      "learning_rate": 1.1322646822786109e-06,
+      "loss": 1.416,
+      "step": 19113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6890954140588978,
+      "learning_rate": 1.1310241094571294e-06,
+      "loss": 0.3032,
+      "step": 19114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0220017218049486,
+      "learning_rate": 1.129784200899528e-06,
+      "loss": 1.0576,
+      "step": 19115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.440563436523452,
+      "learning_rate": 1.1285449566403094e-06,
+      "loss": 1.2812,
+      "step": 19116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.727159093611456,
+      "learning_rate": 1.1273063767139713e-06,
+      "loss": 1.209,
+      "step": 19117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3501248065986435,
+      "learning_rate": 1.1260684611549778e-06,
+      "loss": 1.1367,
+      "step": 19118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.96931365042973,
+      "learning_rate": 1.1248312099977816e-06,
+      "loss": 1.4414,
+      "step": 19119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.817676781403078,
+      "learning_rate": 1.123594623276808e-06,
+      "loss": 1.0928,
+      "step": 19120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.988814569546578,
+      "learning_rate": 1.1223587010264825e-06,
+      "loss": 0.6014,
+      "step": 19121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.362187543341258,
+      "learning_rate": 1.1211234432811857e-06,
+      "loss": 1.9355,
+      "step": 19122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7973969894138793,
+      "learning_rate": 1.1198888500753096e-06,
+      "loss": 1.0996,
+      "step": 19123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.147630005833832,
+      "learning_rate": 1.118654921443199e-06,
+      "loss": 0.813,
+      "step": 19124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.034024488852312,
+      "learning_rate": 1.1174216574192042e-06,
+      "loss": 1.1074,
+      "step": 19125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.428334616556916,
+      "learning_rate": 1.116189058037645e-06,
+      "loss": 1.1895,
+      "step": 19126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1640775934787837,
+      "learning_rate": 1.1149571233328189e-06,
+      "loss": 1.0757,
+      "step": 19127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.123259097600945,
+      "learning_rate": 1.1137258533390066e-06,
+      "loss": 1.0322,
+      "step": 19128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2925470799527865,
+      "learning_rate": 1.1124952480904865e-06,
+      "loss": 1.0322,
+      "step": 19129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9410127779343536,
+      "learning_rate": 1.111265307621498e-06,
+      "loss": 0.5586,
+      "step": 19130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6097064692863374,
+      "learning_rate": 1.110036031966266e-06,
+      "loss": 0.5522,
+      "step": 19131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6481235540614905,
+      "learning_rate": 1.108807421159011e-06,
+      "loss": 1.6152,
+      "step": 19132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1742838941294,
+      "learning_rate": 1.1075794752339164e-06,
+      "loss": 0.7598,
+      "step": 19133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.21535632868281,
+      "learning_rate": 1.1063521942251603e-06,
+      "loss": 1.7988,
+      "step": 19134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4753754957091014,
+      "learning_rate": 1.1051255781668962e-06,
+      "loss": 0.8621,
+      "step": 19135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.938404674417676,
+      "learning_rate": 1.1038996270932579e-06,
+      "loss": 1.085,
+      "step": 19136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9188794779189173,
+      "learning_rate": 1.102674341038365e-06,
+      "loss": 1.126,
+      "step": 19137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3807148056810523,
+      "learning_rate": 1.1014497200363156e-06,
+      "loss": 0.9019,
+      "step": 19138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5496609963121648,
+      "learning_rate": 1.1002257641211932e-06,
+      "loss": 1.0977,
+      "step": 19139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0191582255640745,
+      "learning_rate": 1.0990024733270572e-06,
+      "loss": 1.7734,
+      "step": 19140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.400180824751087,
+      "learning_rate": 1.097779847687949e-06,
+      "loss": 1.0918,
+      "step": 19141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.047425756540672,
+      "learning_rate": 1.0965578872379034e-06,
+      "loss": 0.6296,
+      "step": 19142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9683381512696263,
+      "learning_rate": 1.0953365920109177e-06,
+      "loss": 0.637,
+      "step": 19143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0309185051018748,
+      "learning_rate": 1.094115962040984e-06,
+      "loss": 0.6267,
+      "step": 19144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.589311426435254,
+      "learning_rate": 1.09289599736207e-06,
+      "loss": 1.4197,
+      "step": 19145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7662225282514177,
+      "learning_rate": 1.0916766980081288e-06,
+      "loss": 1.582,
+      "step": 19146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8985385106009067,
+      "learning_rate": 1.0904580640130919e-06,
+      "loss": 1.6797,
+      "step": 19147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.933076815515218,
+      "learning_rate": 1.0892400954108734e-06,
+      "loss": 1.0039,
+      "step": 19148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1871725654773466,
+      "learning_rate": 1.0880227922353685e-06,
+      "loss": 1.1116,
+      "step": 19149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.859467121671661,
+      "learning_rate": 1.0868061545204505e-06,
+      "loss": 1.2129,
+      "step": 19150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.312071580121962,
+      "learning_rate": 1.0855901822999892e-06,
+      "loss": 1.1025,
+      "step": 19151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9554513316154205,
+      "learning_rate": 1.084374875607816e-06,
+      "loss": 0.7871,
+      "step": 19152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4370046085241306,
+      "learning_rate": 1.0831602344777537e-06,
+      "loss": 0.9702,
+      "step": 19153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.014148093551211,
+      "learning_rate": 1.0819462589436003e-06,
+      "loss": 1.043,
+      "step": 19154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.679006595684617,
+      "learning_rate": 1.080732949039151e-06,
+      "loss": 1.1875,
+      "step": 19155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2919823053485833,
+      "learning_rate": 1.0795203047981678e-06,
+      "loss": 0.9443,
+      "step": 19156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.472261659495753,
+      "learning_rate": 1.078308326254393e-06,
+      "loss": 0.7305,
+      "step": 19157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9463751618800105,
+      "learning_rate": 1.0770970134415582e-06,
+      "loss": 0.6836,
+      "step": 19158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.136733478725516,
+      "learning_rate": 1.0758863663933748e-06,
+      "loss": 0.9761,
+      "step": 19159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2451885995001404,
+      "learning_rate": 1.074676385143536e-06,
+      "loss": 0.8594,
+      "step": 19160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8325838406745145,
+      "learning_rate": 1.0734670697257144e-06,
+      "loss": 0.9248,
+      "step": 19161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2127703830140932,
+      "learning_rate": 1.0722584201735581e-06,
+      "loss": 0.96,
+      "step": 19162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3261856091507864,
+      "learning_rate": 1.0710504365207152e-06,
+      "loss": 1.084,
+      "step": 19163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8280277715405637,
+      "learning_rate": 1.0698431188007952e-06,
+      "loss": 1.2617,
+      "step": 19164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.4173945402505375,
+      "learning_rate": 1.0686364670474013e-06,
+      "loss": 0.9397,
+      "step": 19165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1889171068166617,
+      "learning_rate": 1.067430481294107e-06,
+      "loss": 0.9092,
+      "step": 19166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9009103877755271,
+      "learning_rate": 1.066225161574483e-06,
+      "loss": 0.578,
+      "step": 19167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8451568141955343,
+      "learning_rate": 1.0650205079220654e-06,
+      "loss": 1.3262,
+      "step": 19168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.485784358804883,
+      "learning_rate": 1.0638165203703892e-06,
+      "loss": 0.7412,
+      "step": 19169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2040902520944843,
+      "learning_rate": 1.0626131989529525e-06,
+      "loss": 0.5378,
+      "step": 19170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.74206084060412,
+      "learning_rate": 1.0614105437032423e-06,
+      "loss": 0.8987,
+      "step": 19171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.207811173899604,
+      "learning_rate": 1.0602085546547346e-06,
+      "loss": 1.1875,
+      "step": 19172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.090536432954708,
+      "learning_rate": 1.0590072318408778e-06,
+      "loss": 0.6621,
+      "step": 19173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.918017120164688,
+      "learning_rate": 1.0578065752951032e-06,
+      "loss": 1.0918,
+      "step": 19174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.665130848490407,
+      "learning_rate": 1.0566065850508205e-06,
+      "loss": 0.8682,
+      "step": 19175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.880697921827159,
+      "learning_rate": 1.0554072611414334e-06,
+      "loss": 1.0801,
+      "step": 19176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0894718017428042,
+      "learning_rate": 1.054208603600315e-06,
+      "loss": 1.1309,
+      "step": 19177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3640291842216414,
+      "learning_rate": 1.053010612460817e-06,
+      "loss": 0.6406,
+      "step": 19178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1604190273388504,
+      "learning_rate": 1.051813287756287e-06,
+      "loss": 0.8994,
+      "step": 19179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.530862094455063,
+      "learning_rate": 1.0506166295200459e-06,
+      "loss": 1.0527,
+      "step": 19180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.363676502072651,
+      "learning_rate": 1.0494206377853949e-06,
+      "loss": 0.835,
+      "step": 19181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.7789437307617515,
+      "learning_rate": 1.0482253125856184e-06,
+      "loss": 0.9995,
+      "step": 19182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8568199441355744,
+      "learning_rate": 1.047030653953976e-06,
+      "loss": 1.3867,
+      "step": 19183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3389186654368834,
+      "learning_rate": 1.0458366619237215e-06,
+      "loss": 1.3691,
+      "step": 19184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.634465370574836,
+      "learning_rate": 1.044643336528084e-06,
+      "loss": 0.6738,
+      "step": 19185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.946507704691877,
+      "learning_rate": 1.0434506778002674e-06,
+      "loss": 1.1064,
+      "step": 19186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.909185588173774,
+      "learning_rate": 1.0422586857734644e-06,
+      "loss": 0.5049,
+      "step": 19187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.364050369584553,
+      "learning_rate": 1.041067360480849e-06,
+      "loss": 0.9326,
+      "step": 19188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.25513126049801,
+      "learning_rate": 1.039876701955575e-06,
+      "loss": 0.5774,
+      "step": 19189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.659261275660322,
+      "learning_rate": 1.0386867102307824e-06,
+      "loss": 1.6255,
+      "step": 19190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0230635379697284,
+      "learning_rate": 1.0374973853395808e-06,
+      "loss": 0.7812,
+      "step": 19191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.573758884967226,
+      "learning_rate": 1.0363087273150718e-06,
+      "loss": 1.0098,
+      "step": 19192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.8027767769192975,
+      "learning_rate": 1.035120736190337e-06,
+      "loss": 1.2632,
+      "step": 19193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.457981612616082,
+      "learning_rate": 1.0339334119984362e-06,
+      "loss": 0.7394,
+      "step": 19194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.341682436005893,
+      "learning_rate": 1.0327467547724124e-06,
+      "loss": 1.1729,
+      "step": 19195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.242107642476257,
+      "learning_rate": 1.0315607645452836e-06,
+      "loss": 1.0449,
+      "step": 19196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5723533991941037,
+      "learning_rate": 1.0303754413500648e-06,
+      "loss": 1.2334,
+      "step": 19197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0695005732127925,
+      "learning_rate": 1.0291907852197385e-06,
+      "loss": 0.8809,
+      "step": 19198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0387964769371507,
+      "learning_rate": 1.028006796187278e-06,
+      "loss": 0.6504,
+      "step": 19199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.612729885089506,
+      "learning_rate": 1.0268234742856236e-06,
+      "loss": 1.1172,
+      "step": 19200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3137710789607,
+      "learning_rate": 1.0256408195477185e-06,
+      "loss": 0.8262,
+      "step": 19201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.031137732216925,
+      "learning_rate": 1.024458832006467e-06,
+      "loss": 0.8174,
+      "step": 19202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3252469905796476,
+      "learning_rate": 1.0232775116947647e-06,
+      "loss": 1.0273,
+      "step": 19203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2407094090947757,
+      "learning_rate": 1.0220968586454882e-06,
+      "loss": 1.1729,
+      "step": 19204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.018298020167474,
+      "learning_rate": 1.020916872891492e-06,
+      "loss": 1.0811,
+      "step": 19205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.482184446011907,
+      "learning_rate": 1.0197375544656219e-06,
+      "loss": 0.7461,
+      "step": 19206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.907719938271779,
+      "learning_rate": 1.0185589034006931e-06,
+      "loss": 0.8877,
+      "step": 19207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.000054120285205,
+      "learning_rate": 1.0173809197295075e-06,
+      "loss": 1.3765,
+      "step": 19208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9611718825115747,
+      "learning_rate": 1.0162036034848444e-06,
+      "loss": 1.3945,
+      "step": 19209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2759660591817257,
+      "learning_rate": 1.0150269546994745e-06,
+      "loss": 1.2686,
+      "step": 19210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.85187595046257,
+      "learning_rate": 1.0138509734061414e-06,
+      "loss": 0.8804,
+      "step": 19211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8040174421071993,
+      "learning_rate": 1.0126756596375686e-06,
+      "loss": 0.5322,
+      "step": 19212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 27.284257825670394,
+      "learning_rate": 1.0115010134264635e-06,
+      "loss": 2.0581,
+      "step": 19213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.552340963348712,
+      "learning_rate": 1.0103270348055248e-06,
+      "loss": 0.7886,
+      "step": 19214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.2099059312056415,
+      "learning_rate": 1.0091537238074127e-06,
+      "loss": 0.9297,
+      "step": 19215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7793212374854575,
+      "learning_rate": 1.0079810804647927e-06,
+      "loss": 0.7195,
+      "step": 19216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.332965299555101,
+      "learning_rate": 1.0068091048102857e-06,
+      "loss": 1.2598,
+      "step": 19217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.846380688408365,
+      "learning_rate": 1.0056377968765184e-06,
+      "loss": 1.188,
+      "step": 19218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.291513435768445,
+      "learning_rate": 1.0044671566960817e-06,
+      "loss": 0.7061,
+      "step": 19219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.21841022394751,
+      "learning_rate": 1.0032971843015576e-06,
+      "loss": 1.1445,
+      "step": 19220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.262740726204542,
+      "learning_rate": 1.002127879725498e-06,
+      "loss": 0.6357,
+      "step": 19221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.09516898042231,
+      "learning_rate": 1.0009592430004543e-06,
+      "loss": 1.0957,
+      "step": 19222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2248992597083648,
+      "learning_rate": 9.997912741589456e-07,
+      "loss": 0.7996,
+      "step": 19223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.410234413256211,
+      "learning_rate": 9.986239732334701e-07,
+      "loss": 1.2949,
+      "step": 19224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4718946878104924,
+      "learning_rate": 9.974573402565246e-07,
+      "loss": 1.4961,
+      "step": 19225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.969515951367251,
+      "learning_rate": 9.962913752605635e-07,
+      "loss": 1.0762,
+      "step": 19226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1061585947006685,
+      "learning_rate": 9.95126078278047e-07,
+      "loss": 0.8281,
+      "step": 19227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8571274893211145,
+      "learning_rate": 9.93961449341399e-07,
+      "loss": 1.4668,
+      "step": 19228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.21446906278328,
+      "learning_rate": 9.9279748848303e-07,
+      "loss": 1.2129,
+      "step": 19229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.03975995876772,
+      "learning_rate": 9.91634195735333e-07,
+      "loss": 0.8447,
+      "step": 19230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.967697703896134,
+      "learning_rate": 9.904715711306828e-07,
+      "loss": 1.3047,
+      "step": 19231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.88389256152167,
+      "learning_rate": 9.893096147014364e-07,
+      "loss": 0.7764,
+      "step": 19232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.112372519885518,
+      "learning_rate": 9.881483264799234e-07,
+      "loss": 1.1113,
+      "step": 19233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6498972602937467,
+      "learning_rate": 9.869877064984712e-07,
+      "loss": 0.7915,
+      "step": 19234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.343436766991995,
+      "learning_rate": 9.85827754789373e-07,
+      "loss": 0.8643,
+      "step": 19235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3809875781314256,
+      "learning_rate": 9.846684713849142e-07,
+      "loss": 0.9355,
+      "step": 19236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.000003973640353,
+      "learning_rate": 9.835098563173578e-07,
+      "loss": 1.1523,
+      "step": 19237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.243155634161652,
+      "learning_rate": 9.823519096189393e-07,
+      "loss": 0.8232,
+      "step": 19238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.066495869467664,
+      "learning_rate": 9.811946313218938e-07,
+      "loss": 2.3652,
+      "step": 19239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9060652479288427,
+      "learning_rate": 9.800380214584237e-07,
+      "loss": 1.1543,
+      "step": 19240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.361611548825979,
+      "learning_rate": 9.788820800607169e-07,
+      "loss": 1.0293,
+      "step": 19241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.06659597637668,
+      "learning_rate": 9.777268071609391e-07,
+      "loss": 1.168,
+      "step": 19242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7068645211269966,
+      "learning_rate": 9.765722027912455e-07,
+      "loss": 1.3066,
+      "step": 19243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.174384882971093,
+      "learning_rate": 9.754182669837686e-07,
+      "loss": 1.2148,
+      "step": 19244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.1865965274752535,
+      "learning_rate": 9.742649997706216e-07,
+      "loss": 1.0586,
+      "step": 19245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9683919590324876,
+      "learning_rate": 9.73112401183901e-07,
+      "loss": 1.1113,
+      "step": 19246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.272867767152962,
+      "learning_rate": 9.719604712556757e-07,
+      "loss": 0.5479,
+      "step": 19247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7818384894829484,
+      "learning_rate": 9.708092100180143e-07,
+      "loss": 0.8071,
+      "step": 19248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.440715645516603,
+      "learning_rate": 9.696586175029471e-07,
+      "loss": 1.125,
+      "step": 19249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.099677916475601,
+      "learning_rate": 9.685086937425008e-07,
+      "loss": 1.3877,
+      "step": 19250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.45834948510651,
+      "learning_rate": 9.673594387686696e-07,
+      "loss": 1.4326,
+      "step": 19251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.29301895514888,
+      "learning_rate": 9.662108526134417e-07,
+      "loss": 1.4463,
+      "step": 19252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0763842312615135,
+      "learning_rate": 9.65062935308783e-07,
+      "loss": 0.7676,
+      "step": 19253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3433773507460343,
+      "learning_rate": 9.639156868866406e-07,
+      "loss": 1.332,
+      "step": 19254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0556341391082187,
+      "learning_rate": 9.627691073789386e-07,
+      "loss": 1.1338,
+      "step": 19255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.611855670202432,
+      "learning_rate": 9.61623196817585e-07,
+      "loss": 1.0254,
+      "step": 19256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.0113707097079825,
+      "learning_rate": 9.604779552344735e-07,
+      "loss": 1.0327,
+      "step": 19257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.135300863330515,
+      "learning_rate": 9.59333382661473e-07,
+      "loss": 1.1123,
+      "step": 19258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5701023004261447,
+      "learning_rate": 9.581894791304363e-07,
+      "loss": 0.7959,
+      "step": 19259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1536451109242996,
+      "learning_rate": 9.570462446732014e-07,
+      "loss": 0.8706,
+      "step": 19260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.982274182460197,
+      "learning_rate": 9.559036793215792e-07,
+      "loss": 0.6733,
+      "step": 19261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.005169148908606,
+      "learning_rate": 9.54761783107372e-07,
+      "loss": 1.0195,
+      "step": 19262
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1929575822992273,
+      "learning_rate": 9.53620556062354e-07,
+      "loss": 0.876,
+      "step": 19263
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3499309418805416,
+      "learning_rate": 9.524799982182836e-07,
+      "loss": 1.7031,
+      "step": 19264
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9865187917563323,
+      "learning_rate": 9.513401096069102e-07,
+      "loss": 0.4695,
+      "step": 19265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.778045852522055,
+      "learning_rate": 9.502008902599501e-07,
+      "loss": 1.8213,
+      "step": 19266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.19211358379499,
+      "learning_rate": 9.490623402091086e-07,
+      "loss": 1.0566,
+      "step": 19267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7240818401707836,
+      "learning_rate": 9.479244594860686e-07,
+      "loss": 1.5332,
+      "step": 19268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.933232751629567,
+      "learning_rate": 9.46787248122502e-07,
+      "loss": 1.0156,
+      "step": 19269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1445285204762996,
+      "learning_rate": 9.456507061500531e-07,
+      "loss": 0.7666,
+      "step": 19270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.213360317683022,
+      "learning_rate": 9.445148336003518e-07,
+      "loss": 0.5464,
+      "step": 19271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.550347853557891,
+      "learning_rate": 9.433796305050119e-07,
+      "loss": 1.0703,
+      "step": 19272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2727063238739,
+      "learning_rate": 9.42245096895622e-07,
+      "loss": 1.3184,
+      "step": 19273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.101708465114685,
+      "learning_rate": 9.411112328037569e-07,
+      "loss": 0.5635,
+      "step": 19274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.638199725146753,
+      "learning_rate": 9.399780382609746e-07,
+      "loss": 0.7422,
+      "step": 19275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.154474826909137,
+      "learning_rate": 9.388455132988055e-07,
+      "loss": 1.3535,
+      "step": 19276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.034832754287554,
+      "learning_rate": 9.377136579487688e-07,
+      "loss": 0.6018,
+      "step": 19277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.508461261389166,
+      "learning_rate": 9.365824722423699e-07,
+      "loss": 0.6504,
+      "step": 19278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.963129243096534,
+      "learning_rate": 9.354519562110808e-07,
+      "loss": 1.0977,
+      "step": 19279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.286466684144419,
+      "learning_rate": 9.34322109886368e-07,
+      "loss": 0.5752,
+      "step": 19280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2656821408136456,
+      "learning_rate": 9.331929332996731e-07,
+      "loss": 0.3839,
+      "step": 19281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.067656919158848,
+      "learning_rate": 9.320644264824235e-07,
+      "loss": 1.1816,
+      "step": 19282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.206091078863639,
+      "learning_rate": 9.309365894660222e-07,
+      "loss": 0.6738,
+      "step": 19283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.419401544561566,
+      "learning_rate": 9.298094222818577e-07,
+      "loss": 1.0068,
+      "step": 19284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.365048015806682,
+      "learning_rate": 9.286829249612966e-07,
+      "loss": 0.7947,
+      "step": 19285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.510230371680478,
+      "learning_rate": 9.275570975356917e-07,
+      "loss": 0.772,
+      "step": 19286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.514329754533496,
+      "learning_rate": 9.264319400363736e-07,
+      "loss": 0.8687,
+      "step": 19287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.294898190968661,
+      "learning_rate": 9.253074524946531e-07,
+      "loss": 1.7529,
+      "step": 19288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1262962651139077,
+      "learning_rate": 9.24183634941822e-07,
+      "loss": 1.0186,
+      "step": 19289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.251394828341658,
+      "learning_rate": 9.230604874091609e-07,
+      "loss": 0.9521,
+      "step": 19290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2620503605802567,
+      "learning_rate": 9.219380099279252e-07,
+      "loss": 1.0156,
+      "step": 19291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.280022215384468,
+      "learning_rate": 9.208162025293537e-07,
+      "loss": 1.0947,
+      "step": 19292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4480660142724084,
+      "learning_rate": 9.196950652446662e-07,
+      "loss": 1.4414,
+      "step": 19293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.347889557123164,
+      "learning_rate": 9.18574598105057e-07,
+      "loss": 0.9534,
+      "step": 19294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.713717979660266,
+      "learning_rate": 9.174548011417178e-07,
+      "loss": 1.5703,
+      "step": 19295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.663160623241134,
+      "learning_rate": 9.16335674385807e-07,
+      "loss": 1.2207,
+      "step": 19296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.057656330843744,
+      "learning_rate": 9.152172178684693e-07,
+      "loss": 1.0381,
+      "step": 19297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0551714309135174,
+      "learning_rate": 9.14099431620824e-07,
+      "loss": 0.8696,
+      "step": 19298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5210561942798493,
+      "learning_rate": 9.129823156739936e-07,
+      "loss": 0.6011,
+      "step": 19299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4306963331839815,
+      "learning_rate": 9.118658700590616e-07,
+      "loss": 1.5703,
+      "step": 19300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5294344949288057,
+      "learning_rate": 9.107500948070947e-07,
+      "loss": 0.9443,
+      "step": 19301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4938612910703597,
+      "learning_rate": 9.096349899491402e-07,
+      "loss": 1.2397,
+      "step": 19302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.968063235093624,
+      "learning_rate": 9.08520555516243e-07,
+      "loss": 1.5039,
+      "step": 19303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7531562379067243,
+      "learning_rate": 9.074067915394113e-07,
+      "loss": 0.9717,
+      "step": 19304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.398236592479883,
+      "learning_rate": 9.062936980496373e-07,
+      "loss": 0.723,
+      "step": 19305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.45293320528142,
+      "learning_rate": 9.051812750779015e-07,
+      "loss": 1.123,
+      "step": 19306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9496457226625017,
+      "learning_rate": 9.040695226551599e-07,
+      "loss": 1.291,
+      "step": 19307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3568731521117687,
+      "learning_rate": 9.02958440812357e-07,
+      "loss": 0.8506,
+      "step": 19308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.044363228750291,
+      "learning_rate": 9.018480295804127e-07,
+      "loss": 1.1797,
+      "step": 19309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.44109372999744,
+      "learning_rate": 9.007382889902271e-07,
+      "loss": 1.4688,
+      "step": 19310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5337223176803674,
+      "learning_rate": 8.996292190726785e-07,
+      "loss": 0.7212,
+      "step": 19311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.209338253422479,
+      "learning_rate": 8.98520819858642e-07,
+      "loss": 1.2617,
+      "step": 19312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8094682000752984,
+      "learning_rate": 8.974130913789597e-07,
+      "loss": 0.6343,
+      "step": 19313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0794687157470073,
+      "learning_rate": 8.963060336644596e-07,
+      "loss": 1.0732,
+      "step": 19314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7763649547673928,
+      "learning_rate": 8.951996467459451e-07,
+      "loss": 1.0938,
+      "step": 19315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6239937261355224,
+      "learning_rate": 8.940939306542162e-07,
+      "loss": 1.394,
+      "step": 19316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.932666199054911,
+      "learning_rate": 8.929888854200346e-07,
+      "loss": 1.0322,
+      "step": 19317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4361865099271696,
+      "learning_rate": 8.918845110741619e-07,
+      "loss": 1.0156,
+      "step": 19318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.622568093898614,
+      "learning_rate": 8.907808076473261e-07,
+      "loss": 1.4629,
+      "step": 19319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.72867071881388,
+      "learning_rate": 8.896777751702501e-07,
+      "loss": 0.8298,
+      "step": 19320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5059719414699841,
+      "learning_rate": 8.885754136736257e-07,
+      "loss": 0.6493,
+      "step": 19321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1641128347655254,
+      "learning_rate": 8.874737231881314e-07,
+      "loss": 0.6279,
+      "step": 19322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4485004351231674,
+      "learning_rate": 8.86372703744423e-07,
+      "loss": 1.1299,
+      "step": 19323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.645776015585265,
+      "learning_rate": 8.852723553731484e-07,
+      "loss": 1.2695,
+      "step": 19324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9443830647658733,
+      "learning_rate": 8.841726781049304e-07,
+      "loss": 0.8789,
+      "step": 19325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9052694984216996,
+      "learning_rate": 8.830736719703637e-07,
+      "loss": 0.9053,
+      "step": 19326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9138501730726187,
+      "learning_rate": 8.819753370000405e-07,
+      "loss": 1.1309,
+      "step": 19327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.804111092079562,
+      "learning_rate": 8.808776732245255e-07,
+      "loss": 0.9971,
+      "step": 19328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.878319221652548,
+      "learning_rate": 8.797806806743664e-07,
+      "loss": 1.1909,
+      "step": 19329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7816137645112635,
+      "learning_rate": 8.786843593800914e-07,
+      "loss": 0.8184,
+      "step": 19330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8092796538640474,
+      "learning_rate": 8.775887093722124e-07,
+      "loss": 0.8174,
+      "step": 19331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9499332441005164,
+      "learning_rate": 8.76493730681216e-07,
+      "loss": 0.8167,
+      "step": 19332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9960243585103092,
+      "learning_rate": 8.753994233375806e-07,
+      "loss": 0.9893,
+      "step": 19333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.933980090915608,
+      "learning_rate": 8.743057873717598e-07,
+      "loss": 0.8745,
+      "step": 19334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.160633781953469,
+      "learning_rate": 8.732128228141817e-07,
+      "loss": 0.9922,
+      "step": 19335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6310010162490745,
+      "learning_rate": 8.721205296952723e-07,
+      "loss": 1.1797,
+      "step": 19336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8292036216438055,
+      "learning_rate": 8.710289080454264e-07,
+      "loss": 1.166,
+      "step": 19337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.473660570930907,
+      "learning_rate": 8.699379578950228e-07,
+      "loss": 0.6484,
+      "step": 19338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.609080159730707,
+      "learning_rate": 8.688476792744232e-07,
+      "loss": 1.5801,
+      "step": 19339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.10414330077257,
+      "learning_rate": 8.677580722139672e-07,
+      "loss": 0.5654,
+      "step": 19340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.41685277540516,
+      "learning_rate": 8.666691367439833e-07,
+      "loss": 1.0801,
+      "step": 19341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2381858073019347,
+      "learning_rate": 8.655808728947695e-07,
+      "loss": 1.4492,
+      "step": 19342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.428786843857491,
+      "learning_rate": 8.644932806966155e-07,
+      "loss": 0.603,
+      "step": 19343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.982002848094915,
+      "learning_rate": 8.63406360179786e-07,
+      "loss": 0.8198,
+      "step": 19344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.615975519402972,
+      "learning_rate": 8.623201113745316e-07,
+      "loss": 1.2295,
+      "step": 19345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3642354252734994,
+      "learning_rate": 8.612345343110867e-07,
+      "loss": 0.7717,
+      "step": 19346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0959223395294546,
+      "learning_rate": 8.601496290196548e-07,
+      "loss": 0.9287,
+      "step": 19347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.841567157326813,
+      "learning_rate": 8.59065395530434e-07,
+      "loss": 1.8535,
+      "step": 19348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9889725826635734,
+      "learning_rate": 8.57981833873589e-07,
+      "loss": 1.2217,
+      "step": 19349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7520664879086962,
+      "learning_rate": 8.568989440792874e-07,
+      "loss": 1.0488,
+      "step": 19350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.62844149956178,
+      "learning_rate": 8.558167261776579e-07,
+      "loss": 1.418,
+      "step": 19351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9865340254601513,
+      "learning_rate": 8.547351801988207e-07,
+      "loss": 1.0171,
+      "step": 19352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.493662098141316,
+      "learning_rate": 8.536543061728713e-07,
+      "loss": 1.1895,
+      "step": 19353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1539869599106036,
+      "learning_rate": 8.525741041298912e-07,
+      "loss": 0.7104,
+      "step": 19354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4226319145449193,
+      "learning_rate": 8.51494574099948e-07,
+      "loss": 0.7932,
+      "step": 19355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.686488027557669,
+      "learning_rate": 8.504157161130788e-07,
+      "loss": 1.3281,
+      "step": 19356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.00931746399285,
+      "learning_rate": 8.493375301993067e-07,
+      "loss": 1.877,
+      "step": 19357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.626533060629373,
+      "learning_rate": 8.482600163886412e-07,
+      "loss": 0.9316,
+      "step": 19358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6693540319572353,
+      "learning_rate": 8.471831747110664e-07,
+      "loss": 1.1377,
+      "step": 19359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5043411710323604,
+      "learning_rate": 8.461070051965531e-07,
+      "loss": 1.7461,
+      "step": 19360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7633121970840477,
+      "learning_rate": 8.450315078750464e-07,
+      "loss": 0.8372,
+      "step": 19361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.912892615578641,
+      "learning_rate": 8.439566827764811e-07,
+      "loss": 1.1758,
+      "step": 19362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5476951155122385,
+      "learning_rate": 8.428825299307636e-07,
+      "loss": 1.2578,
+      "step": 19363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7149863665328016,
+      "learning_rate": 8.418090493677949e-07,
+      "loss": 0.9697,
+      "step": 19364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2069380232053404,
+      "learning_rate": 8.407362411174458e-07,
+      "loss": 1.6797,
+      "step": 19365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.487846683899374,
+      "learning_rate": 8.396641052095699e-07,
+      "loss": 1.4336,
+      "step": 19366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.078715001603135,
+      "learning_rate": 8.385926416740075e-07,
+      "loss": 1.1836,
+      "step": 19367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.264361583697603,
+      "learning_rate": 8.375218505405763e-07,
+      "loss": 0.832,
+      "step": 19368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9072780651008248,
+      "learning_rate": 8.364517318390746e-07,
+      "loss": 0.6731,
+      "step": 19369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.995918897204137,
+      "learning_rate": 8.353822855992844e-07,
+      "loss": 0.9551,
+      "step": 19370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.434806900984258,
+      "learning_rate": 8.343135118509676e-07,
+      "loss": 0.8721,
+      "step": 19371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7109217534146244,
+      "learning_rate": 8.332454106238646e-07,
+      "loss": 1.0938,
+      "step": 19372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.036757145147744,
+      "learning_rate": 8.321779819477099e-07,
+      "loss": 0.8906,
+      "step": 19373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.5022339185536655,
+      "learning_rate": 8.311112258521964e-07,
+      "loss": 1.1211,
+      "step": 19374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7916255682202205,
+      "learning_rate": 8.300451423670225e-07,
+      "loss": 1.1738,
+      "step": 19375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8499721324963616,
+      "learning_rate": 8.289797315218534e-07,
+      "loss": 0.7905,
+      "step": 19376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5197058328867374,
+      "learning_rate": 8.279149933463376e-07,
+      "loss": 1.3145,
+      "step": 19377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.701393058871227,
+      "learning_rate": 8.26850927870107e-07,
+      "loss": 1.0371,
+      "step": 19378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4529549120597682,
+      "learning_rate": 8.257875351227739e-07,
+      "loss": 0.4229,
+      "step": 19379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.565676139665703,
+      "learning_rate": 8.247248151339343e-07,
+      "loss": 1.0195,
+      "step": 19380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8297771875669673,
+      "learning_rate": 8.23662767933156e-07,
+      "loss": 0.803,
+      "step": 19381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5063989761481946,
+      "learning_rate": 8.226013935500043e-07,
+      "loss": 1.1797,
+      "step": 19382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.145351454567979,
+      "learning_rate": 8.215406920140112e-07,
+      "loss": 0.8794,
+      "step": 19383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.096466089214245,
+      "learning_rate": 8.204806633547002e-07,
+      "loss": 0.7542,
+      "step": 19384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.495438992824056,
+      "learning_rate": 8.194213076015699e-07,
+      "loss": 1.3203,
+      "step": 19385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.100681593012665,
+      "learning_rate": 8.183626247840998e-07,
+      "loss": 1.2002,
+      "step": 19386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.156312696381857,
+      "learning_rate": 8.173046149317493e-07,
+      "loss": 1.293,
+      "step": 19387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9741494317239883,
+      "learning_rate": 8.162472780739699e-07,
+      "loss": 0.7153,
+      "step": 19388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.34360312098262,
+      "learning_rate": 8.151906142401855e-07,
+      "loss": 1.1426,
+      "step": 19389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8486100491161561,
+      "learning_rate": 8.141346234598002e-07,
+      "loss": 1.2656,
+      "step": 19390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.795815805379289,
+      "learning_rate": 8.130793057621988e-07,
+      "loss": 1.123,
+      "step": 19391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.975860432807923,
+      "learning_rate": 8.120246611767524e-07,
+      "loss": 0.7883,
+      "step": 19392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.054137353552593,
+      "learning_rate": 8.109706897328179e-07,
+      "loss": 1.0713,
+      "step": 19393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.066759844760357,
+      "learning_rate": 8.099173914597191e-07,
+      "loss": 1.332,
+      "step": 19394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.088255433773213,
+      "learning_rate": 8.088647663867715e-07,
+      "loss": 0.4749,
+      "step": 19395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7132335105185206,
+      "learning_rate": 8.078128145432712e-07,
+      "loss": 1.2227,
+      "step": 19396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4372199460073487,
+      "learning_rate": 8.067615359584918e-07,
+      "loss": 1.3398,
+      "step": 19397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.988254946809689,
+      "learning_rate": 8.057109306616878e-07,
+      "loss": 0.9453,
+      "step": 19398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.39348422530585,
+      "learning_rate": 8.046609986820969e-07,
+      "loss": 0.731,
+      "step": 19399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.793184882917909,
+      "learning_rate": 8.036117400489429e-07,
+      "loss": 1.2734,
+      "step": 19400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.274491539626063,
+      "learning_rate": 8.025631547914247e-07,
+      "loss": 1.1492,
+      "step": 19401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6253045223347815,
+      "learning_rate": 8.015152429387246e-07,
+      "loss": 1.0947,
+      "step": 19402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0882997317197387,
+      "learning_rate": 8.004680045200025e-07,
+      "loss": 0.7368,
+      "step": 19403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.717751944618347,
+      "learning_rate": 7.994214395644017e-07,
+      "loss": 1.248,
+      "step": 19404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.376077233531754,
+      "learning_rate": 7.983755481010519e-07,
+      "loss": 1.0859,
+      "step": 19405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1603646227365947,
+      "learning_rate": 7.973303301590601e-07,
+      "loss": 1.0012,
+      "step": 19406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.704389535215792,
+      "learning_rate": 7.962857857675116e-07,
+      "loss": 1.2734,
+      "step": 19407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.335613583421388,
+      "learning_rate": 7.952419149554746e-07,
+      "loss": 0.8164,
+      "step": 19408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4815595496198206,
+      "learning_rate": 7.941987177519982e-07,
+      "loss": 1.3223,
+      "step": 19409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4032212058306355,
+      "learning_rate": 7.931561941861232e-07,
+      "loss": 0.9468,
+      "step": 19410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.361755887672416,
+      "learning_rate": 7.921143442868568e-07,
+      "loss": 1.2656,
+      "step": 19411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.577815032687377,
+      "learning_rate": 7.910731680831895e-07,
+      "loss": 1.6875,
+      "step": 19412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.169647650892432,
+      "learning_rate": 7.900326656041013e-07,
+      "loss": 0.9253,
+      "step": 19413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3161536166566736,
+      "learning_rate": 7.889928368785493e-07,
+      "loss": 0.7666,
+      "step": 19414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9797075302206644,
+      "learning_rate": 7.879536819354716e-07,
+      "loss": 0.5327,
+      "step": 19415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0133340517046996,
+      "learning_rate": 7.869152008037839e-07,
+      "loss": 1.1426,
+      "step": 19416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.361430202017537,
+      "learning_rate": 7.858773935123881e-07,
+      "loss": 0.873,
+      "step": 19417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2900567613255296,
+      "learning_rate": 7.848402600901667e-07,
+      "loss": 0.7881,
+      "step": 19418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8896956315848703,
+      "learning_rate": 7.838038005659853e-07,
+      "loss": 1.6582,
+      "step": 19419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.258748418854156,
+      "learning_rate": 7.827680149686878e-07,
+      "loss": 0.8613,
+      "step": 19420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.342813514368427,
+      "learning_rate": 7.817329033270926e-07,
+      "loss": 1.333,
+      "step": 19421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.196897864188841,
+      "learning_rate": 7.806984656700156e-07,
+      "loss": 1.3594,
+      "step": 19422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.140754668886491,
+      "learning_rate": 7.796647020262393e-07,
+      "loss": 1.541,
+      "step": 19423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3583008803256584,
+      "learning_rate": 7.786316124245324e-07,
+      "loss": 0.9492,
+      "step": 19424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.178303268777748,
+      "learning_rate": 7.775991968936469e-07,
+      "loss": 0.9727,
+      "step": 19425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1555417874258125,
+      "learning_rate": 7.765674554623181e-07,
+      "loss": 1.252,
+      "step": 19426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.625940792730247,
+      "learning_rate": 7.755363881592536e-07,
+      "loss": 1.3184,
+      "step": 19427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.200044220380421,
+      "learning_rate": 7.7450599501315e-07,
+      "loss": 0.9238,
+      "step": 19428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.283974330650667,
+      "learning_rate": 7.734762760526814e-07,
+      "loss": 1.6875,
+      "step": 19429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.558953587098117,
+      "learning_rate": 7.724472313065057e-07,
+      "loss": 0.8474,
+      "step": 19430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.088328273713279,
+      "learning_rate": 7.714188608032608e-07,
+      "loss": 0.7014,
+      "step": 19431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.2659377732474075,
+      "learning_rate": 7.703911645715683e-07,
+      "loss": 0.5769,
+      "step": 19432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9804766953890176,
+      "learning_rate": 7.693641426400222e-07,
+      "loss": 1.1763,
+      "step": 19433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6031158666919887,
+      "learning_rate": 7.683377950372078e-07,
+      "loss": 0.9453,
+      "step": 19434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2255101979831284,
+      "learning_rate": 7.673121217916912e-07,
+      "loss": 0.6519,
+      "step": 19435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.769921663436798,
+      "learning_rate": 7.662871229320106e-07,
+      "loss": 1.1465,
+      "step": 19436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.969192197304307,
+      "learning_rate": 7.652627984866933e-07,
+      "loss": 1.1846,
+      "step": 19437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8264462878881407,
+      "learning_rate": 7.642391484842442e-07,
+      "loss": 0.7327,
+      "step": 19438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6842810404523743,
+      "learning_rate": 7.632161729531572e-07,
+      "loss": 0.9209,
+      "step": 19439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9602404254494,
+      "learning_rate": 7.621938719218985e-07,
+      "loss": 0.7812,
+      "step": 19440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6174286740833643,
+      "learning_rate": 7.611722454189147e-07,
+      "loss": 0.6987,
+      "step": 19441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3081081028022044,
+      "learning_rate": 7.601512934726384e-07,
+      "loss": 0.9863,
+      "step": 19442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.987024415893753,
+      "learning_rate": 7.591310161114862e-07,
+      "loss": 0.6327,
+      "step": 19443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.402769404957241,
+      "learning_rate": 7.581114133638489e-07,
+      "loss": 0.9189,
+      "step": 19444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.675934710324419,
+      "learning_rate": 7.570924852581013e-07,
+      "loss": 1.0049,
+      "step": 19445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.08587588086902,
+      "learning_rate": 7.560742318225983e-07,
+      "loss": 0.7822,
+      "step": 19446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9128862086771274,
+      "learning_rate": 7.550566530856812e-07,
+      "loss": 1.1416,
+      "step": 19447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3522081470964356,
+      "learning_rate": 7.54039749075669e-07,
+      "loss": 0.6777,
+      "step": 19448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.829616497214426,
+      "learning_rate": 7.530235198208585e-07,
+      "loss": 0.8698,
+      "step": 19449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.515317530826633,
+      "learning_rate": 7.520079653495354e-07,
+      "loss": 1.4707,
+      "step": 19450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1704663435675684,
+      "learning_rate": 7.509930856899549e-07,
+      "loss": 1.666,
+      "step": 19451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6720470631643924,
+      "learning_rate": 7.499788808703695e-07,
+      "loss": 1.0508,
+      "step": 19452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5421203979564577,
+      "learning_rate": 7.489653509189981e-07,
+      "loss": 1.2227,
+      "step": 19453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.10512122399228,
+      "learning_rate": 7.47952495864046e-07,
+      "loss": 0.9121,
+      "step": 19454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.307015734406974,
+      "learning_rate": 7.469403157337046e-07,
+      "loss": 1.082,
+      "step": 19455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6792508977894234,
+      "learning_rate": 7.459288105561374e-07,
+      "loss": 1.4453,
+      "step": 19456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0588539726346657,
+      "learning_rate": 7.449179803595024e-07,
+      "loss": 0.9854,
+      "step": 19457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.712066246610246,
+      "learning_rate": 7.439078251719245e-07,
+      "loss": 1.5078,
+      "step": 19458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6985622497057586,
+      "learning_rate": 7.428983450215171e-07,
+      "loss": 0.499,
+      "step": 19459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.310901914455402,
+      "learning_rate": 7.418895399363746e-07,
+      "loss": 1.2285,
+      "step": 19460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.631924623824623,
+      "learning_rate": 7.408814099445688e-07,
+      "loss": 0.5801,
+      "step": 19461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7356450303427664,
+      "learning_rate": 7.398739550741607e-07,
+      "loss": 1.144,
+      "step": 19462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.742684201618245,
+      "learning_rate": 7.388671753531806e-07,
+      "loss": 1.9746,
+      "step": 19463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3109725856632335,
+      "learning_rate": 7.378610708096506e-07,
+      "loss": 0.9775,
+      "step": 19464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.189428084861065,
+      "learning_rate": 7.368556414715705e-07,
+      "loss": 1.0684,
+      "step": 19465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2676291184848205,
+      "learning_rate": 7.358508873669206e-07,
+      "loss": 0.6943,
+      "step": 19466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6786532744423126,
+      "learning_rate": 7.34846808523662e-07,
+      "loss": 1.5176,
+      "step": 19467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.248739841108979,
+      "learning_rate": 7.33843404969739e-07,
+      "loss": 1.4004,
+      "step": 19468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.074059776636577,
+      "learning_rate": 7.328406767330765e-07,
+      "loss": 0.5889,
+      "step": 19469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5680077141619178,
+      "learning_rate": 7.318386238415769e-07,
+      "loss": 1.1504,
+      "step": 19470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2569547825763996,
+      "learning_rate": 7.308372463231322e-07,
+      "loss": 0.7881,
+      "step": 19471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.067224351434802,
+      "learning_rate": 7.298365442056004e-07,
+      "loss": 1.0977,
+      "step": 19472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1174450256705746,
+      "learning_rate": 7.288365175168427e-07,
+      "loss": 1.5723,
+      "step": 19473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7463931221618014,
+      "learning_rate": 7.278371662846811e-07,
+      "loss": 0.8389,
+      "step": 19474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.203995877471502,
+      "learning_rate": 7.268384905369324e-07,
+      "loss": 1.0195,
+      "step": 19475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6692154346702943,
+      "learning_rate": 7.258404903013827e-07,
+      "loss": 1.2539,
+      "step": 19476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.568932792878959,
+      "learning_rate": 7.248431656058125e-07,
+      "loss": 1.2188,
+      "step": 19477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4851429070381648,
+      "learning_rate": 7.238465164779773e-07,
+      "loss": 1.1494,
+      "step": 19478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2829755762692776,
+      "learning_rate": 7.228505429456106e-07,
+      "loss": 0.8805,
+      "step": 19479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0404367310395206,
+      "learning_rate": 7.218552450364263e-07,
+      "loss": 1.0557,
+      "step": 19480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8211479323367517,
+      "learning_rate": 7.208606227781273e-07,
+      "loss": 1.3047,
+      "step": 19481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.015833960247534,
+      "learning_rate": 7.198666761983968e-07,
+      "loss": 1.1084,
+      "step": 19482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.593912004882748,
+      "learning_rate": 7.188734053248881e-07,
+      "loss": 1.2334,
+      "step": 19483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1122663846874308,
+      "learning_rate": 7.178808101852508e-07,
+      "loss": 1.1152,
+      "step": 19484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.248577328590901,
+      "learning_rate": 7.16888890807102e-07,
+      "loss": 0.5654,
+      "step": 19485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0950563824698225,
+      "learning_rate": 7.1589764721805e-07,
+      "loss": 1.042,
+      "step": 19486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.874259090373853,
+      "learning_rate": 7.149070794456842e-07,
+      "loss": 0.9033,
+      "step": 19487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.757383491496467,
+      "learning_rate": 7.139171875175654e-07,
+      "loss": 1.3457,
+      "step": 19488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.417014318255138,
+      "learning_rate": 7.129279714612441e-07,
+      "loss": 1.3008,
+      "step": 19489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7117123774368728,
+      "learning_rate": 7.11939431304251e-07,
+      "loss": 0.7021,
+      "step": 19490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.371470962625189,
+      "learning_rate": 7.109515670740974e-07,
+      "loss": 1.5234,
+      "step": 19491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9408950665434213,
+      "learning_rate": 7.099643787982724e-07,
+      "loss": 0.8669,
+      "step": 19492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6861894509305753,
+      "learning_rate": 7.089778665042457e-07,
+      "loss": 1.0403,
+      "step": 19493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.079983377783373,
+      "learning_rate": 7.079920302194814e-07,
+      "loss": 1.3047,
+      "step": 19494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.896703674020174,
+      "learning_rate": 7.070068699714078e-07,
+      "loss": 0.8677,
+      "step": 19495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2380253777040955,
+      "learning_rate": 7.060223857874443e-07,
+      "loss": 1.248,
+      "step": 19496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.66052357200454,
+      "learning_rate": 7.050385776949859e-07,
+      "loss": 0.8643,
+      "step": 19497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.539298366448428,
+      "learning_rate": 7.040554457214161e-07,
+      "loss": 0.6592,
+      "step": 19498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7054420658431986,
+      "learning_rate": 7.030729898940908e-07,
+      "loss": 1.1836,
+      "step": 19499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0942162056500075,
+      "learning_rate": 7.020912102403548e-07,
+      "loss": 0.6423,
+      "step": 19500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1214144132397648,
+      "learning_rate": 7.011101067875254e-07,
+      "loss": 0.8994,
+      "step": 19501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6871712748240615,
+      "learning_rate": 7.001296795629108e-07,
+      "loss": 0.5967,
+      "step": 19502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0104029231680562,
+      "learning_rate": 6.991499285937953e-07,
+      "loss": 0.6846,
+      "step": 19503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3783747699710027,
+      "learning_rate": 6.981708539074455e-07,
+      "loss": 0.6981,
+      "step": 19504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.301196840747286,
+      "learning_rate": 6.971924555311066e-07,
+      "loss": 1.1992,
+      "step": 19505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.583176167126934,
+      "learning_rate": 6.962147334920066e-07,
+      "loss": 0.5996,
+      "step": 19506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5093174871977433,
+      "learning_rate": 6.952376878173572e-07,
+      "loss": 1.207,
+      "step": 19507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.221664213372346,
+      "learning_rate": 6.942613185343505e-07,
+      "loss": 1.3145,
+      "step": 19508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.29356022383971,
+      "learning_rate": 6.932856256701564e-07,
+      "loss": 1.2676,
+      "step": 19509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.074250014477183,
+      "learning_rate": 6.923106092519255e-07,
+      "loss": 1.2031,
+      "step": 19510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.182741602822836,
+      "learning_rate": 6.913362693067943e-07,
+      "loss": 0.9473,
+      "step": 19511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4594800720648324,
+      "learning_rate": 6.9036260586188e-07,
+      "loss": 0.5532,
+      "step": 19512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.12565131584193,
+      "learning_rate": 6.893896189442778e-07,
+      "loss": 1.1572,
+      "step": 19513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7635277161220326,
+      "learning_rate": 6.884173085810658e-07,
+      "loss": 1.4492,
+      "step": 19514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.0851174693219265,
+      "learning_rate": 6.87445674799303e-07,
+      "loss": 0.6361,
+      "step": 19515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.534649391231163,
+      "learning_rate": 6.864747176260289e-07,
+      "loss": 1.248,
+      "step": 19516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.095086232542195,
+      "learning_rate": 6.85504437088269e-07,
+      "loss": 0.6831,
+      "step": 19517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.224944266146295,
+      "learning_rate": 6.845348332130158e-07,
+      "loss": 1.3481,
+      "step": 19518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9210643143975894,
+      "learning_rate": 6.835659060272642e-07,
+      "loss": 0.8296,
+      "step": 19519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6781813055860915,
+      "learning_rate": 6.825976555579705e-07,
+      "loss": 1.1475,
+      "step": 19520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.214581724676467,
+      "learning_rate": 6.816300818320881e-07,
+      "loss": 1.1309,
+      "step": 19521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.053746402496335,
+      "learning_rate": 6.806631848765399e-07,
+      "loss": 1.3945,
+      "step": 19522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2347710531855944,
+      "learning_rate": 6.796969647182322e-07,
+      "loss": 0.5449,
+      "step": 19523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.638132152089796,
+      "learning_rate": 6.787314213840601e-07,
+      "loss": 1.8066,
+      "step": 19524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.705084813779492,
+      "learning_rate": 6.777665549008883e-07,
+      "loss": 0.876,
+      "step": 19525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2052944675274273,
+      "learning_rate": 6.768023652955758e-07,
+      "loss": 1.0723,
+      "step": 19526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.905812821075097,
+      "learning_rate": 6.758388525949483e-07,
+      "loss": 1.043,
+      "step": 19527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9220406388440514,
+      "learning_rate": 6.748760168258234e-07,
+      "loss": 0.437,
+      "step": 19528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.139528870996005,
+      "learning_rate": 6.739138580149989e-07,
+      "loss": 0.5864,
+      "step": 19529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.687095273329077,
+      "learning_rate": 6.729523761892454e-07,
+      "loss": 1.5879,
+      "step": 19530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9901924465805485,
+      "learning_rate": 6.719915713753217e-07,
+      "loss": 0.5942,
+      "step": 19531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.971546729764388,
+      "learning_rate": 6.710314435999732e-07,
+      "loss": 0.6846,
+      "step": 19532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.646505487086525,
+      "learning_rate": 6.700719928899146e-07,
+      "loss": 0.7295,
+      "step": 19533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.001440742424748,
+      "learning_rate": 6.691132192718469e-07,
+      "loss": 0.8584,
+      "step": 19534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5682130479798886,
+      "learning_rate": 6.681551227724514e-07,
+      "loss": 0.5918,
+      "step": 19535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.844139659301443,
+      "learning_rate": 6.671977034183957e-07,
+      "loss": 1.5469,
+      "step": 19536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.332703960392221,
+      "learning_rate": 6.662409612363224e-07,
+      "loss": 1.2305,
+      "step": 19537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3370492409333092,
+      "learning_rate": 6.652848962528574e-07,
+      "loss": 0.9873,
+      "step": 19538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6433421945443416,
+      "learning_rate": 6.643295084946044e-07,
+      "loss": 1.2344,
+      "step": 19539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5106633699584213,
+      "learning_rate": 6.633747979881533e-07,
+      "loss": 1.0654,
+      "step": 19540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1045170699301674,
+      "learning_rate": 6.624207647600772e-07,
+      "loss": 1.0879,
+      "step": 19541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.806094972965074,
+      "learning_rate": 6.614674088369243e-07,
+      "loss": 0.918,
+      "step": 19542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.55439371425048,
+      "learning_rate": 6.605147302452236e-07,
+      "loss": 0.9463,
+      "step": 19543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5618193117758605,
+      "learning_rate": 6.595627290114869e-07,
+      "loss": 0.9355,
+      "step": 19544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8778781417620307,
+      "learning_rate": 6.586114051622127e-07,
+      "loss": 1.2354,
+      "step": 19545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.598746625780535,
+      "learning_rate": 6.57660758723877e-07,
+      "loss": 1.6836,
+      "step": 19546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.350835339224076,
+      "learning_rate": 6.56710789722928e-07,
+      "loss": 1.0076,
+      "step": 19547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.65104497017339,
+      "learning_rate": 6.557614981858085e-07,
+      "loss": 1.0742,
+      "step": 19548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.680356340536989,
+      "learning_rate": 6.548128841389334e-07,
+      "loss": 0.8503,
+      "step": 19549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6789765151569,
+      "learning_rate": 6.538649476087094e-07,
+      "loss": 1.0029,
+      "step": 19550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.422508599863446,
+      "learning_rate": 6.529176886215127e-07,
+      "loss": 1.2129,
+      "step": 19551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.532397824605283,
+      "learning_rate": 6.519711072037e-07,
+      "loss": 1.082,
+      "step": 19552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1892255712418174,
+      "learning_rate": 6.51025203381625e-07,
+      "loss": 0.4932,
+      "step": 19553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.207375344998491,
+      "learning_rate": 6.500799771816029e-07,
+      "loss": 1.0723,
+      "step": 19554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.445448509065759,
+      "learning_rate": 6.491354286299433e-07,
+      "loss": 1.6035,
+      "step": 19555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1313696984961177,
+      "learning_rate": 6.481915577529307e-07,
+      "loss": 0.8047,
+      "step": 19556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.203778746616027,
+      "learning_rate": 6.47248364576833e-07,
+      "loss": 1.6055,
+      "step": 19557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9706319866062618,
+      "learning_rate": 6.463058491278984e-07,
+      "loss": 1.2988,
+      "step": 19558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8947500830042672,
+      "learning_rate": 6.453640114323617e-07,
+      "loss": 1.2891,
+      "step": 19559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1070758864356858,
+      "learning_rate": 6.444228515164269e-07,
+      "loss": 1.0693,
+      "step": 19560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1619301767357824,
+      "learning_rate": 6.434823694062897e-07,
+      "loss": 1.0244,
+      "step": 19561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.023184620141623,
+      "learning_rate": 6.425425651281208e-07,
+      "loss": 1.208,
+      "step": 19562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7908313515798358,
+      "learning_rate": 6.4160343870808e-07,
+      "loss": 1.124,
+      "step": 19563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7963293822818376,
+      "learning_rate": 6.406649901722989e-07,
+      "loss": 0.5372,
+      "step": 19564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.439023624590696,
+      "learning_rate": 6.3972721954689e-07,
+      "loss": 1.8184,
+      "step": 19565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9933306836054274,
+      "learning_rate": 6.387901268579605e-07,
+      "loss": 1.1855,
+      "step": 19566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.035277689464449,
+      "learning_rate": 6.378537121315781e-07,
+      "loss": 0.7378,
+      "step": 19567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.350459563435332,
+      "learning_rate": 6.369179753938165e-07,
+      "loss": 1.1104,
+      "step": 19568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6227642965549438,
+      "learning_rate": 6.359829166707048e-07,
+      "loss": 0.7817,
+      "step": 19569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.549017574143181,
+      "learning_rate": 6.350485359882696e-07,
+      "loss": 0.7766,
+      "step": 19570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4449660570717855,
+      "learning_rate": 6.341148333725178e-07,
+      "loss": 1.1074,
+      "step": 19571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.579502916577312,
+      "learning_rate": 6.331818088494285e-07,
+      "loss": 0.6289,
+      "step": 19572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.183544615600485,
+      "learning_rate": 6.322494624449698e-07,
+      "loss": 1.0439,
+      "step": 19573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3714311790123683,
+      "learning_rate": 6.313177941850906e-07,
+      "loss": 1.3281,
+      "step": 19574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.152556642502615,
+      "learning_rate": 6.303868040957145e-07,
+      "loss": 0.9727,
+      "step": 19575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2204280471533973,
+      "learning_rate": 6.294564922027513e-07,
+      "loss": 1.0122,
+      "step": 19576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.440034477897068,
+      "learning_rate": 6.285268585320969e-07,
+      "loss": 0.894,
+      "step": 19577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.979192017509997,
+      "learning_rate": 6.275979031096141e-07,
+      "loss": 1.541,
+      "step": 19578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.456181649716853,
+      "learning_rate": 6.266696259611626e-07,
+      "loss": 1.6309,
+      "step": 19579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3268258151385774,
+      "learning_rate": 6.257420271125747e-07,
+      "loss": 0.74,
+      "step": 19580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.2349935137718555,
+      "learning_rate": 6.248151065896602e-07,
+      "loss": 1.3984,
+      "step": 19581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8333149461113145,
+      "learning_rate": 6.238888644182178e-07,
+      "loss": 0.8242,
+      "step": 19582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5676586880907584,
+      "learning_rate": 6.22963300624027e-07,
+      "loss": 1.0146,
+      "step": 19583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.346135667343233,
+      "learning_rate": 6.220384152328423e-07,
+      "loss": 1.2178,
+      "step": 19584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2663809731831694,
+      "learning_rate": 6.211142082704041e-07,
+      "loss": 1.2969,
+      "step": 19585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2979568656872351,
+      "learning_rate": 6.201906797624335e-07,
+      "loss": 0.4711,
+      "step": 19586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4694616523355766,
+      "learning_rate": 6.192678297346294e-07,
+      "loss": 0.5791,
+      "step": 19587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.288863033418184,
+      "learning_rate": 6.183456582126795e-07,
+      "loss": 0.9929,
+      "step": 19588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8674989517101994,
+      "learning_rate": 6.174241652222412e-07,
+      "loss": 1.1562,
+      "step": 19589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.250079744732436,
+      "learning_rate": 6.165033507889661e-07,
+      "loss": 0.9014,
+      "step": 19590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1445958480360163,
+      "learning_rate": 6.155832149384699e-07,
+      "loss": 0.4312,
+      "step": 19591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.857554372034895,
+      "learning_rate": 6.146637576963709e-07,
+      "loss": 0.79,
+      "step": 19592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2591633171226504,
+      "learning_rate": 6.137449790882516e-07,
+      "loss": 1.2422,
+      "step": 19593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.151485947238875,
+      "learning_rate": 6.128268791396802e-07,
+      "loss": 0.9785,
+      "step": 19594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0729276667683587,
+      "learning_rate": 6.119094578762086e-07,
+      "loss": 1.2578,
+      "step": 19595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.508118227567638,
+      "learning_rate": 6.109927153233691e-07,
+      "loss": 1.1523,
+      "step": 19596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7779261714209507,
+      "learning_rate": 6.100766515066747e-07,
+      "loss": 0.2308,
+      "step": 19597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.236767534955774,
+      "learning_rate": 6.091612664516188e-07,
+      "loss": 0.8359,
+      "step": 19598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6629196803335002,
+      "learning_rate": 6.082465601836729e-07,
+      "loss": 1.3359,
+      "step": 19599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.092777115847784,
+      "learning_rate": 6.07332532728297e-07,
+      "loss": 0.8262,
+      "step": 19600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.594498308082333,
+      "learning_rate": 6.064191841109262e-07,
+      "loss": 1.3027,
+      "step": 19601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.040646837474269,
+      "learning_rate": 6.055065143569794e-07,
+      "loss": 1.3555,
+      "step": 19602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9907606228789385,
+      "learning_rate": 6.045945234918526e-07,
+      "loss": 1.5059,
+      "step": 19603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.569634716331957,
+      "learning_rate": 6.036832115409286e-07,
+      "loss": 0.3662,
+      "step": 19604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9859295260361587,
+      "learning_rate": 6.02772578529573e-07,
+      "loss": 1.2432,
+      "step": 19605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1638490958126018,
+      "learning_rate": 6.018626244831211e-07,
+      "loss": 0.9814,
+      "step": 19606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.537986081174616,
+      "learning_rate": 6.009533494269026e-07,
+      "loss": 1.2686,
+      "step": 19607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.584399639044475,
+      "learning_rate": 6.00044753386217e-07,
+      "loss": 1.0879,
+      "step": 19608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.7814057200311915,
+      "learning_rate": 5.991368363863547e-07,
+      "loss": 1.6621,
+      "step": 19609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.692904921602275,
+      "learning_rate": 5.98229598452582e-07,
+      "loss": 1.2402,
+      "step": 19610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8600868214587014,
+      "learning_rate": 5.973230396101398e-07,
+      "loss": 1.166,
+      "step": 19611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.062365937704633,
+      "learning_rate": 5.964171598842688e-07,
+      "loss": 1.1836,
+      "step": 19612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.30043154480543,
+      "learning_rate": 5.955119593001685e-07,
+      "loss": 1.0928,
+      "step": 19613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.62384075052538,
+      "learning_rate": 5.946074378830407e-07,
+      "loss": 1.3389,
+      "step": 19614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9045668503537572,
+      "learning_rate": 5.937035956580489e-07,
+      "loss": 0.3552,
+      "step": 19615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.292199182773142,
+      "learning_rate": 5.928004326503506e-07,
+      "loss": 0.8779,
+      "step": 19616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.142887117539523,
+      "learning_rate": 5.918979488850812e-07,
+      "loss": 1.0459,
+      "step": 19617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7085518577643533,
+      "learning_rate": 5.909961443873569e-07,
+      "loss": 0.5918,
+      "step": 19618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.684158554499141,
+      "learning_rate": 5.900950191822713e-07,
+      "loss": 1.2812,
+      "step": 19619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8183435752829626,
+      "learning_rate": 5.891945732949017e-07,
+      "loss": 1.2568,
+      "step": 19620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.891744928239507,
+      "learning_rate": 5.882948067503114e-07,
+      "loss": 1.2422,
+      "step": 19621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5688445470631964,
+      "learning_rate": 5.873957195735385e-07,
+      "loss": 1.0449,
+      "step": 19622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9868238697516896,
+      "learning_rate": 5.86497311789605e-07,
+      "loss": 0.9333,
+      "step": 19623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.690564116565584,
+      "learning_rate": 5.855995834235128e-07,
+      "loss": 0.833,
+      "step": 19624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.360587307540951,
+      "learning_rate": 5.847025345002422e-07,
+      "loss": 0.9375,
+      "step": 19625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.84131819487697,
+      "learning_rate": 5.838061650447646e-07,
+      "loss": 1.1602,
+      "step": 19626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8929046714417184,
+      "learning_rate": 5.829104750820213e-07,
+      "loss": 0.8291,
+      "step": 19627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6665670753796498,
+      "learning_rate": 5.820154646369369e-07,
+      "loss": 0.9316,
+      "step": 19628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2565149884024702,
+      "learning_rate": 5.811211337344219e-07,
+      "loss": 0.3521,
+      "step": 19629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.047601857661973,
+      "learning_rate": 5.802274823993647e-07,
+      "loss": 0.9014,
+      "step": 19630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2272157908263712,
+      "learning_rate": 5.793345106566372e-07,
+      "loss": 1.0869,
+      "step": 19631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2115566074303077,
+      "learning_rate": 5.784422185310861e-07,
+      "loss": 1.0513,
+      "step": 19632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.068042311554246,
+      "learning_rate": 5.775506060475472e-07,
+      "loss": 1.1992,
+      "step": 19633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5632352937056786,
+      "learning_rate": 5.766596732308338e-07,
+      "loss": 1.188,
+      "step": 19634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.031677417488823,
+      "learning_rate": 5.757694201057373e-07,
+      "loss": 1.252,
+      "step": 19635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.058246851039863,
+      "learning_rate": 5.748798466970379e-07,
+      "loss": 0.8502,
+      "step": 19636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.071353405965183,
+      "learning_rate": 5.739909530294824e-07,
+      "loss": 1.1086,
+      "step": 19637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.06521139231319,
+      "learning_rate": 5.731027391278204e-07,
+      "loss": 0.6228,
+      "step": 19638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.328977142008338,
+      "learning_rate": 5.722152050167656e-07,
+      "loss": 3.2715,
+      "step": 19639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9375226419417941,
+      "learning_rate": 5.713283507210148e-07,
+      "loss": 0.6498,
+      "step": 19640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6067639838912156,
+      "learning_rate": 5.704421762652484e-07,
+      "loss": 0.9316,
+      "step": 19641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5128892413467159,
+      "learning_rate": 5.695566816741327e-07,
+      "loss": 0.4247,
+      "step": 19642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2181446162925993,
+      "learning_rate": 5.686718669723118e-07,
+      "loss": 1.6152,
+      "step": 19643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8224801932620736,
+      "learning_rate": 5.67787732184405e-07,
+      "loss": 0.9395,
+      "step": 19644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5376687317921687,
+      "learning_rate": 5.669042773350202e-07,
+      "loss": 1.1523,
+      "step": 19645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9746259166051083,
+      "learning_rate": 5.660215024487408e-07,
+      "loss": 1.3406,
+      "step": 19646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.039485355872406,
+      "learning_rate": 5.651394075501359e-07,
+      "loss": 1.0391,
+      "step": 19647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8822758699808433,
+      "learning_rate": 5.642579926637554e-07,
+      "loss": 0.9363,
+      "step": 19648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.629854834622989,
+      "learning_rate": 5.633772578141266e-07,
+      "loss": 0.8525,
+      "step": 19649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0443351919366988,
+      "learning_rate": 5.62497203025758e-07,
+      "loss": 1.0859,
+      "step": 19650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.500575285524401,
+      "learning_rate": 5.616178283231466e-07,
+      "loss": 0.9912,
+      "step": 19651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6096280826112412,
+      "learning_rate": 5.607391337307615e-07,
+      "loss": 1.1562,
+      "step": 19652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.7833446110478945,
+      "learning_rate": 5.598611192730586e-07,
+      "loss": 1.0898,
+      "step": 19653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.177962601525055,
+      "learning_rate": 5.589837849744678e-07,
+      "loss": 0.422,
+      "step": 19654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.279508370047948,
+      "learning_rate": 5.581071308594116e-07,
+      "loss": 0.8638,
+      "step": 19655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.121461727475898,
+      "learning_rate": 5.572311569522842e-07,
+      "loss": 1.2363,
+      "step": 19656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5524024192995873,
+      "learning_rate": 5.563558632774607e-07,
+      "loss": 1.1934,
+      "step": 19657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1021344282319125,
+      "learning_rate": 5.554812498593021e-07,
+      "loss": 1.5723,
+      "step": 19658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.250625550149564,
+      "learning_rate": 5.546073167221499e-07,
+      "loss": 0.7891,
+      "step": 19659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.109161676344747,
+      "learning_rate": 5.537340638903237e-07,
+      "loss": 1.125,
+      "step": 19660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6690348202952023,
+      "learning_rate": 5.528614913881291e-07,
+      "loss": 0.9629,
+      "step": 19661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.366051228069434,
+      "learning_rate": 5.519895992398439e-07,
+      "loss": 0.8853,
+      "step": 19662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5776518796274406,
+      "learning_rate": 5.511183874697345e-07,
+      "loss": 1.0127,
+      "step": 19663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1777359077208733,
+      "learning_rate": 5.502478561020513e-07,
+      "loss": 1.0938,
+      "step": 19664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.442550578862547,
+      "learning_rate": 5.493780051610137e-07,
+      "loss": 1.5684,
+      "step": 19665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.097891966577444,
+      "learning_rate": 5.485088346708356e-07,
+      "loss": 1.0801,
+      "step": 19666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.993507371499984,
+      "learning_rate": 5.476403446556977e-07,
+      "loss": 1.1855,
+      "step": 19667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9692679505511808,
+      "learning_rate": 5.467725351397779e-07,
+      "loss": 1.1387,
+      "step": 19668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.674576670581961,
+      "learning_rate": 5.459054061472207e-07,
+      "loss": 0.9033,
+      "step": 19669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7176577462627147,
+      "learning_rate": 5.450389577021625e-07,
+      "loss": 1.6328,
+      "step": 19670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.4751552253881215,
+      "learning_rate": 5.441731898287144e-07,
+      "loss": 1.3574,
+      "step": 19671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.6496141499229875,
+      "learning_rate": 5.433081025509712e-07,
+      "loss": 1.0156,
+      "step": 19672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.223379564897217,
+      "learning_rate": 5.424436958930051e-07,
+      "loss": 0.9072,
+      "step": 19673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8085761779320344,
+      "learning_rate": 5.415799698788776e-07,
+      "loss": 0.6758,
+      "step": 19674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.309875737692121,
+      "learning_rate": 5.407169245326193e-07,
+      "loss": 0.6548,
+      "step": 19675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8563672911710156,
+      "learning_rate": 5.398545598782528e-07,
+      "loss": 1.4727,
+      "step": 19676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7052688509324705,
+      "learning_rate": 5.389928759397783e-07,
+      "loss": 1.0508,
+      "step": 19677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2209481110056997,
+      "learning_rate": 5.381318727411683e-07,
+      "loss": 0.8369,
+      "step": 19678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.085597182738359,
+      "learning_rate": 5.372715503063952e-07,
+      "loss": 1.2676,
+      "step": 19679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0821952437754416,
+      "learning_rate": 5.364119086593928e-07,
+      "loss": 1.498,
+      "step": 19680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.034230172970664,
+      "learning_rate": 5.355529478240889e-07,
+      "loss": 0.9204,
+      "step": 19681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1150101730426676,
+      "learning_rate": 5.346946678243897e-07,
+      "loss": 0.8447,
+      "step": 19682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.480084727439678,
+      "learning_rate": 5.338370686841759e-07,
+      "loss": 1.1426,
+      "step": 19683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.761250974125779,
+      "learning_rate": 5.329801504273147e-07,
+      "loss": 1.2266,
+      "step": 19684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1173198137780553,
+      "learning_rate": 5.321239130776589e-07,
+      "loss": 1.292,
+      "step": 19685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.682176338899966,
+      "learning_rate": 5.312683566590315e-07,
+      "loss": 0.8467,
+      "step": 19686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.515330017674139,
+      "learning_rate": 5.304134811952438e-07,
+      "loss": 0.9619,
+      "step": 19687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8798255495539955,
+      "learning_rate": 5.295592867100852e-07,
+      "loss": 0.553,
+      "step": 19688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6609944917522257,
+      "learning_rate": 5.287057732273337e-07,
+      "loss": 0.9192,
+      "step": 19689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.784222161307906,
+      "learning_rate": 5.2785294077074e-07,
+      "loss": 0.603,
+      "step": 19690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.819347614844828,
+      "learning_rate": 5.27000789364035e-07,
+      "loss": 0.7197,
+      "step": 19691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3251686527241073,
+      "learning_rate": 5.261493190309303e-07,
+      "loss": 1.166,
+      "step": 19692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1330631755129628,
+      "learning_rate": 5.252985297951318e-07,
+      "loss": 1.6191,
+      "step": 19693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4943512041280216,
+      "learning_rate": 5.244484216803098e-07,
+      "loss": 1.0645,
+      "step": 19694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.27108614647134,
+      "learning_rate": 5.235989947101255e-07,
+      "loss": 1.123,
+      "step": 19695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.279749688028279,
+      "learning_rate": 5.227502489082159e-07,
+      "loss": 1.8125,
+      "step": 19696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.263795330970277,
+      "learning_rate": 5.219021842981981e-07,
+      "loss": 1.3145,
+      "step": 19697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.916595421783782,
+      "learning_rate": 5.210548009036836e-07,
+      "loss": 1.1104,
+      "step": 19698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.162025511678572,
+      "learning_rate": 5.202080987482455e-07,
+      "loss": 0.916,
+      "step": 19699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.347924024806851,
+      "learning_rate": 5.193620778554536e-07,
+      "loss": 0.7523,
+      "step": 19700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.13907050338305,
+      "learning_rate": 5.185167382488449e-07,
+      "loss": 1.104,
+      "step": 19701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9695606076671441,
+      "learning_rate": 5.176720799519502e-07,
+      "loss": 1.2432,
+      "step": 19702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8337593809348234,
+      "learning_rate": 5.16828102988276e-07,
+      "loss": 0.8159,
+      "step": 19703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1563693486678583,
+      "learning_rate": 5.15984807381309e-07,
+      "loss": 0.8613,
+      "step": 19704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.554733695916183,
+      "learning_rate": 5.151421931545164e-07,
+      "loss": 1.4648,
+      "step": 19705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.731493315859684,
+      "learning_rate": 5.143002603313462e-07,
+      "loss": 1.4336,
+      "step": 19706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8634817467387403,
+      "learning_rate": 5.134590089352353e-07,
+      "loss": 0.7327,
+      "step": 19707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0321870256470618,
+      "learning_rate": 5.126184389895899e-07,
+      "loss": 0.894,
+      "step": 19708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.05434380324899,
+      "learning_rate": 5.117785505178024e-07,
+      "loss": 1.4492,
+      "step": 19709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9056794610378303,
+      "learning_rate": 5.109393435432542e-07,
+      "loss": 1.0869,
+      "step": 19710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.390132678830217,
+      "learning_rate": 5.101008180892902e-07,
+      "loss": 0.791,
+      "step": 19711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0477755363734045,
+      "learning_rate": 5.092629741792531e-07,
+      "loss": 1.0422,
+      "step": 19712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3809623327946685,
+      "learning_rate": 5.084258118364549e-07,
+      "loss": 1.0615,
+      "step": 19713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0339007174887063,
+      "learning_rate": 5.07589331084199e-07,
+      "loss": 1.1621,
+      "step": 19714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4408946729640744,
+      "learning_rate": 5.067535319457584e-07,
+      "loss": 0.7031,
+      "step": 19715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.538904182311874,
+      "learning_rate": 5.059184144443952e-07,
+      "loss": 0.8682,
+      "step": 19716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9139166992366703,
+      "learning_rate": 5.050839786033545e-07,
+      "loss": 1.2207,
+      "step": 19717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.723747833657246,
+      "learning_rate": 5.042502244458513e-07,
+      "loss": 0.7173,
+      "step": 19718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.747714299747805,
+      "learning_rate": 5.034171519950947e-07,
+      "loss": 0.8613,
+      "step": 19719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0886323929738655,
+      "learning_rate": 5.025847612742663e-07,
+      "loss": 1.0938,
+      "step": 19720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.912653875705519,
+      "learning_rate": 5.017530523065306e-07,
+      "loss": 0.2859,
+      "step": 19721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8150562409221345,
+      "learning_rate": 5.009220251150304e-07,
+      "loss": 1.1885,
+      "step": 19722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7633133330529946,
+      "learning_rate": 5.000916797228999e-07,
+      "loss": 0.9062,
+      "step": 19723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4474929984034453,
+      "learning_rate": 4.99262016153243e-07,
+      "loss": 0.7366,
+      "step": 19724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.114257928398705,
+      "learning_rate": 4.984330344291521e-07,
+      "loss": 1.4775,
+      "step": 19725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.987846552328896,
+      "learning_rate": 4.97604734573695e-07,
+      "loss": 1.2627,
+      "step": 19726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.549335852808623,
+      "learning_rate": 4.9677711660992e-07,
+      "loss": 1.2598,
+      "step": 19727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7477252395120964,
+      "learning_rate": 4.959501805608668e-07,
+      "loss": 1.2178,
+      "step": 19728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.862378139979187,
+      "learning_rate": 4.951239264495422e-07,
+      "loss": 1.9453,
+      "step": 19729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9708711611106164,
+      "learning_rate": 4.942983542989416e-07,
+      "loss": 0.9521,
+      "step": 19730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.580854053668383,
+      "learning_rate": 4.934734641320438e-07,
+      "loss": 1.6973,
+      "step": 19731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4995573605638217,
+      "learning_rate": 4.926492559718027e-07,
+      "loss": 1.1553,
+      "step": 19732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3375908999976422,
+      "learning_rate": 4.918257298411555e-07,
+      "loss": 1.4785,
+      "step": 19733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.508580190847197,
+      "learning_rate": 4.9100288576302e-07,
+      "loss": 1.5527,
+      "step": 19734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.571167499292437,
+      "learning_rate": 4.901807237602946e-07,
+      "loss": 1.1499,
+      "step": 19735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.423598309690526,
+      "learning_rate": 4.893592438558664e-07,
+      "loss": 1.6621,
+      "step": 19736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9037781735127837,
+      "learning_rate": 4.885384460725894e-07,
+      "loss": 1.0254,
+      "step": 19737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.395096032401425,
+      "learning_rate": 4.877183304333093e-07,
+      "loss": 0.9097,
+      "step": 19738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.005461952010758,
+      "learning_rate": 4.868988969608496e-07,
+      "loss": 1.2002,
+      "step": 19739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2961817103797597,
+      "learning_rate": 4.860801456780139e-07,
+      "loss": 0.2773,
+      "step": 19740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4926586124378125,
+      "learning_rate": 4.8526207660759e-07,
+      "loss": 0.874,
+      "step": 19741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.7013455057707105,
+      "learning_rate": 4.844446897723398e-07,
+      "loss": 0.7517,
+      "step": 19742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6868202769316385,
+      "learning_rate": 4.836279851950148e-07,
+      "loss": 0.8872,
+      "step": 19743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1233739802703413,
+      "learning_rate": 4.828119628983413e-07,
+      "loss": 0.8174,
+      "step": 19744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3628870858240174,
+      "learning_rate": 4.819966229050315e-07,
+      "loss": 1.5234,
+      "step": 19745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.313380821118055,
+      "learning_rate": 4.81181965237773e-07,
+      "loss": 1.3965,
+      "step": 19746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5544780962960427,
+      "learning_rate": 4.803679899192392e-07,
+      "loss": 1.0801,
+      "step": 19747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.983013672713649,
+      "learning_rate": 4.795546969720816e-07,
+      "loss": 1.0684,
+      "step": 19748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6795546716256213,
+      "learning_rate": 4.787420864189374e-07,
+      "loss": 1.2422,
+      "step": 19749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.353047949520608,
+      "learning_rate": 4.779301582824164e-07,
+      "loss": 1.3379,
+      "step": 19750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.268340594550775,
+      "learning_rate": 4.771189125851144e-07,
+      "loss": 0.9102,
+      "step": 19751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.350085237471972,
+      "learning_rate": 4.763083493496107e-07,
+      "loss": 1.2354,
+      "step": 19752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.041736004748395,
+      "learning_rate": 4.754984685984648e-07,
+      "loss": 1.209,
+      "step": 19753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.532808443611379,
+      "learning_rate": 4.7468927035420887e-07,
+      "loss": 0.728,
+      "step": 19754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2502753830974043,
+      "learning_rate": 4.738807546393692e-07,
+      "loss": 0.8882,
+      "step": 19755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0028934830134597,
+      "learning_rate": 4.730729214764418e-07,
+      "loss": 0.6924,
+      "step": 19756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.522097659000144,
+      "learning_rate": 4.7226577088791133e-07,
+      "loss": 1.0273,
+      "step": 19757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.810730780336539,
+      "learning_rate": 4.714593028962405e-07,
+      "loss": 0.8936,
+      "step": 19758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.666995763527872,
+      "learning_rate": 4.7065351752386965e-07,
+      "loss": 0.9854,
+      "step": 19759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.58250898297522,
+      "learning_rate": 4.698484147932253e-07,
+      "loss": 1.1035,
+      "step": 19760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9749567301192035,
+      "learning_rate": 4.6904399472671166e-07,
+      "loss": 1.2393,
+      "step": 19761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7320805977958966,
+      "learning_rate": 4.682402573467193e-07,
+      "loss": 1.2949,
+      "step": 19762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.714258276291606,
+      "learning_rate": 4.674372026756135e-07,
+      "loss": 1.293,
+      "step": 19763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.152039088921776,
+      "learning_rate": 4.6663483073574324e-07,
+      "loss": 1.1016,
+      "step": 19764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8511717822688043,
+      "learning_rate": 4.658331415494377e-07,
+      "loss": 1.2441,
+      "step": 19765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7978293392976057,
+      "learning_rate": 4.650321351390097e-07,
+      "loss": 0.9336,
+      "step": 19766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.518430900650698,
+      "learning_rate": 4.642318115267469e-07,
+      "loss": 1.2236,
+      "step": 19767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3284795094367268,
+      "learning_rate": 4.6343217073492596e-07,
+      "loss": 1.0293,
+      "step": 19768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.924052360462458,
+      "learning_rate": 4.626332127857985e-07,
+      "loss": 0.8975,
+      "step": 19769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0910940832921514,
+      "learning_rate": 4.618349377015996e-07,
+      "loss": 1.2271,
+      "step": 19770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6561700247898368,
+      "learning_rate": 4.610373455045419e-07,
+      "loss": 1.2441,
+      "step": 19771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.912270493530818,
+      "learning_rate": 4.6024043621682724e-07,
+      "loss": 0.7446,
+      "step": 19772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.625259931501408,
+      "learning_rate": 4.594442098606322e-07,
+      "loss": 1.0645,
+      "step": 19773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.911428943865576,
+      "learning_rate": 4.586486664581141e-07,
+      "loss": 1.0889,
+      "step": 19774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0306770103418246,
+      "learning_rate": 4.578538060314108e-07,
+      "loss": 1.1523,
+      "step": 19775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.151714032058358,
+      "learning_rate": 4.570596286026491e-07,
+      "loss": 1.4609,
+      "step": 19776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.144044614120821,
+      "learning_rate": 4.5626613419392226e-07,
+      "loss": 0.707,
+      "step": 19777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.857066602576047,
+      "learning_rate": 4.554733228273184e-07,
+      "loss": 1.2344,
+      "step": 19778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.778527030562094,
+      "learning_rate": 4.546811945249002e-07,
+      "loss": 1.001,
+      "step": 19779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.674532409422675,
+      "learning_rate": 4.538897493087113e-07,
+      "loss": 0.8384,
+      "step": 19780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4127335178140805,
+      "learning_rate": 4.5309898720078125e-07,
+      "loss": 0.777,
+      "step": 19781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.408241933740417,
+      "learning_rate": 4.5230890822310913e-07,
+      "loss": 1.2031,
+      "step": 19782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.207333363510488,
+      "learning_rate": 4.515195123976884e-07,
+      "loss": 1.3516,
+      "step": 19783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8484118470393294,
+      "learning_rate": 4.507307997464849e-07,
+      "loss": 0.8025,
+      "step": 19784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5117778860104494,
+      "learning_rate": 4.4994277029145324e-07,
+      "loss": 1.1572,
+      "step": 19785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.06874940993194,
+      "learning_rate": 4.4915542405451483e-07,
+      "loss": 0.6777,
+      "step": 19786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3028650351899245,
+      "learning_rate": 4.4836876105758817e-07,
+      "loss": 1.0996,
+      "step": 19787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4995995200777075,
+      "learning_rate": 4.4758278132256414e-07,
+      "loss": 0.9141,
+      "step": 19788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9372910943698587,
+      "learning_rate": 4.467974848713169e-07,
+      "loss": 1.1074,
+      "step": 19789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.101399113841336,
+      "learning_rate": 4.4601287172569563e-07,
+      "loss": 0.853,
+      "step": 19790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.659913404096917,
+      "learning_rate": 4.452289419075439e-07,
+      "loss": 1.2734,
+      "step": 19791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.48564678982194,
+      "learning_rate": 4.4444569543867496e-07,
+      "loss": 1.0,
+      "step": 19792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.623784910855551,
+      "learning_rate": 4.4366313234088796e-07,
+      "loss": 0.8721,
+      "step": 19793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.337133624079401,
+      "learning_rate": 4.4288125263595434e-07,
+      "loss": 0.7676,
+      "step": 19794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8763330105934517,
+      "learning_rate": 4.421000563456401e-07,
+      "loss": 0.8765,
+      "step": 19795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.837885920290319,
+      "learning_rate": 4.4131954349168613e-07,
+      "loss": 1.7207,
+      "step": 19796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.514496335875397,
+      "learning_rate": 4.405397140958084e-07,
+      "loss": 1.29,
+      "step": 19797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.347515022467392,
+      "learning_rate": 4.3976056817971176e-07,
+      "loss": 0.9326,
+      "step": 19798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.893605869461233,
+      "learning_rate": 4.389821057650789e-07,
+      "loss": 0.9287,
+      "step": 19799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.105212137582831,
+      "learning_rate": 4.382043268735786e-07,
+      "loss": 1.4238,
+      "step": 19800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5461789501398966,
+      "learning_rate": 4.3742723152685183e-07,
+      "loss": 0.7792,
+      "step": 19801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8668284178351393,
+      "learning_rate": 4.3665081974652856e-07,
+      "loss": 1.084,
+      "step": 19802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.910898145665152,
+      "learning_rate": 4.3587509155420814e-07,
+      "loss": 1.3711,
+      "step": 19803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.632605893402452,
+      "learning_rate": 4.351000469714872e-07,
+      "loss": 1.1152,
+      "step": 19804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.296033996054867,
+      "learning_rate": 4.3432568601993187e-07,
+      "loss": 1.4727,
+      "step": 19805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6785441921076723,
+      "learning_rate": 4.335520087210887e-07,
+      "loss": 0.8562,
+      "step": 19806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6748252775939436,
+      "learning_rate": 4.327790150964933e-07,
+      "loss": 0.8184,
+      "step": 19807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9958241498997344,
+      "learning_rate": 4.320067051676563e-07,
+      "loss": 1.5469,
+      "step": 19808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9090355058659427,
+      "learning_rate": 4.312350789560743e-07,
+      "loss": 1.4102,
+      "step": 19809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.895746097119769,
+      "learning_rate": 4.304641364832163e-07,
+      "loss": 1.0686,
+      "step": 19810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.393035552758235,
+      "learning_rate": 4.2969387777053726e-07,
+      "loss": 1.0146,
+      "step": 19811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8642097894588956,
+      "learning_rate": 4.289243028394757e-07,
+      "loss": 0.6152,
+      "step": 19812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7318378483323027,
+      "learning_rate": 4.281554117114506e-07,
+      "loss": 0.7124,
+      "step": 19813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0231939338581073,
+      "learning_rate": 4.2738720440785583e-07,
+      "loss": 1.4395,
+      "step": 19814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.248737823710093,
+      "learning_rate": 4.266196809500689e-07,
+      "loss": 0.5278,
+      "step": 19815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.357892887697586,
+      "learning_rate": 4.258528413594559e-07,
+      "loss": 1.3789,
+      "step": 19816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.257686202268827,
+      "learning_rate": 4.2508668565734987e-07,
+      "loss": 0.3926,
+      "step": 19817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8451755848957228,
+      "learning_rate": 4.243212138650809e-07,
+      "loss": 0.9424,
+      "step": 19818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.388095738577151,
+      "learning_rate": 4.2355642600394873e-07,
+      "loss": 0.9707,
+      "step": 19819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.324352305445873,
+      "learning_rate": 4.2279232209523344e-07,
+      "loss": 0.9805,
+      "step": 19820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9444953903227398,
+      "learning_rate": 4.220289021602042e-07,
+      "loss": 1.2734,
+      "step": 19821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.496783616530045,
+      "learning_rate": 4.212661662201023e-07,
+      "loss": 1.1387,
+      "step": 19822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.639996315855997,
+      "learning_rate": 4.2050411429616076e-07,
+      "loss": 1.084,
+      "step": 19823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4477205454944095,
+      "learning_rate": 4.197427464095793e-07,
+      "loss": 0.5386,
+      "step": 19824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3972885908966433,
+      "learning_rate": 4.1898206258155215e-07,
+      "loss": 1.2148,
+      "step": 19825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8214985663057308,
+      "learning_rate": 4.182220628332484e-07,
+      "loss": 1.2783,
+      "step": 19826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3615295475700977,
+      "learning_rate": 4.174627471858178e-07,
+      "loss": 1.2725,
+      "step": 19827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1854262244357883,
+      "learning_rate": 4.1670411566038793e-07,
+      "loss": 0.9375,
+      "step": 19828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.447388081938478,
+      "learning_rate": 4.15946168278078e-07,
+      "loss": 1.0059,
+      "step": 19829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.572843656134658,
+      "learning_rate": 4.1518890505997666e-07,
+      "loss": 0.8667,
+      "step": 19830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.616462841165372,
+      "learning_rate": 4.1443232602716155e-07,
+      "loss": 0.9688,
+      "step": 19831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.096801639923631,
+      "learning_rate": 4.1367643120068244e-07,
+      "loss": 1.2715,
+      "step": 19832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1780519536164853,
+      "learning_rate": 4.129212206015809e-07,
+      "loss": 0.7891,
+      "step": 19833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8243772005873278,
+      "learning_rate": 4.121666942508734e-07,
+      "loss": 0.832,
+      "step": 19834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4390379137734244,
+      "learning_rate": 4.1141285216955707e-07,
+      "loss": 1.3535,
+      "step": 19835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.01532169396138,
+      "learning_rate": 4.1065969437860954e-07,
+      "loss": 0.9717,
+      "step": 19836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.536690765667879,
+      "learning_rate": 4.099072208989918e-07,
+      "loss": 1.3047,
+      "step": 19837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.126538141927478,
+      "learning_rate": 4.091554317516483e-07,
+      "loss": 1.4121,
+      "step": 19838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.15757856431272,
+      "learning_rate": 4.084043269574983e-07,
+      "loss": 1.123,
+      "step": 19839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4674157508759844,
+      "learning_rate": 4.0765390653744183e-07,
+      "loss": 1.123,
+      "step": 19840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.988626896795318,
+      "learning_rate": 4.0690417051236775e-07,
+      "loss": 0.9941,
+      "step": 19841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.948525712647986,
+      "learning_rate": 4.061551189031371e-07,
+      "loss": 0.8721,
+      "step": 19842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6097713104259417,
+      "learning_rate": 4.0540675173059985e-07,
+      "loss": 0.9795,
+      "step": 19843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.003914827478421,
+      "learning_rate": 4.0465906901557824e-07,
+      "loss": 1.4844,
+      "step": 19844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1955072174057886,
+      "learning_rate": 4.0391207077887794e-07,
+      "loss": 0.7158,
+      "step": 19845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9537533950429276,
+      "learning_rate": 4.0316575704129613e-07,
+      "loss": 0.5686,
+      "step": 19846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.814161211721222,
+      "learning_rate": 4.024201278235995e-07,
+      "loss": 1.8066,
+      "step": 19847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.53053518492503,
+      "learning_rate": 4.016751831465354e-07,
+      "loss": 0.6777,
+      "step": 19848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6240355425390667,
+      "learning_rate": 4.0093092303083723e-07,
+      "loss": 0.7617,
+      "step": 19849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1235937387644754,
+      "learning_rate": 4.001873474972162e-07,
+      "loss": 0.9351,
+      "step": 19850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.31013749559754,
+      "learning_rate": 3.994444565663669e-07,
+      "loss": 1.2969,
+      "step": 19851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.927132536211404,
+      "learning_rate": 3.9870225025896434e-07,
+      "loss": 1.0986,
+      "step": 19852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.512766570568242,
+      "learning_rate": 3.9796072859566157e-07,
+      "loss": 0.9424,
+      "step": 19853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.35307770516472,
+      "learning_rate": 3.972198915970976e-07,
+      "loss": 0.8574,
+      "step": 19854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.387992531282368,
+      "learning_rate": 3.964797392838865e-07,
+      "loss": 1.1768,
+      "step": 19855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.905353835778859,
+      "learning_rate": 3.957402716766312e-07,
+      "loss": 0.7588,
+      "step": 19856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.151742065294505,
+      "learning_rate": 3.950014887959069e-07,
+      "loss": 1.1074,
+      "step": 19857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1745646681418154,
+      "learning_rate": 3.9426339066227224e-07,
+      "loss": 0.7925,
+      "step": 19858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.217947586458448,
+      "learning_rate": 3.935259772962718e-07,
+      "loss": 1.6074,
+      "step": 19859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.510414366239599,
+      "learning_rate": 3.927892487184254e-07,
+      "loss": 0.7344,
+      "step": 19860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1094616907460875,
+      "learning_rate": 3.920532049492387e-07,
+      "loss": 1.0625,
+      "step": 19861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.7723006160052375,
+      "learning_rate": 3.9131784600918984e-07,
+      "loss": 1.3906,
+      "step": 19862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.641057600946764,
+      "learning_rate": 3.905831719187486e-07,
+      "loss": 0.8994,
+      "step": 19863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.868211111345894,
+      "learning_rate": 3.8984918269835967e-07,
+      "loss": 1.1201,
+      "step": 19864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2658859628801586,
+      "learning_rate": 3.891158783684512e-07,
+      "loss": 0.5552,
+      "step": 19865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.504572558516126,
+      "learning_rate": 3.883832589494263e-07,
+      "loss": 1.168,
+      "step": 19866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.6973237229544775,
+      "learning_rate": 3.8765132446167694e-07,
+      "loss": 1.1426,
+      "step": 19867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1143359586092085,
+      "learning_rate": 3.8692007492557024e-07,
+      "loss": 1.1895,
+      "step": 19868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.851518074851956,
+      "learning_rate": 3.8618951036145936e-07,
+      "loss": 1.2656,
+      "step": 19869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.978323417972449,
+      "learning_rate": 3.854596307896724e-07,
+      "loss": 1.2115,
+      "step": 19870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.6553179941031635,
+      "learning_rate": 3.847304362305265e-07,
+      "loss": 1.127,
+      "step": 19871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.440159349553754,
+      "learning_rate": 3.840019267043082e-07,
+      "loss": 1.1299,
+      "step": 19872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4797869527211183,
+      "learning_rate": 3.8327410223129576e-07,
+      "loss": 0.6011,
+      "step": 19873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0082410065677387,
+      "learning_rate": 3.825469628317452e-07,
+      "loss": 0.9512,
+      "step": 19874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9567660153894626,
+      "learning_rate": 3.8182050852589026e-07,
+      "loss": 1.1001,
+      "step": 19875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.497081865471172,
+      "learning_rate": 3.810947393339481e-07,
+      "loss": 0.7139,
+      "step": 19876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.613103705819114,
+      "learning_rate": 3.8036965527611657e-07,
+      "loss": 0.9199,
+      "step": 19877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4503400109853306,
+      "learning_rate": 3.7964525637257664e-07,
+      "loss": 0.8965,
+      "step": 19878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.719675542328899,
+      "learning_rate": 3.7892154264348446e-07,
+      "loss": 1.4297,
+      "step": 19879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.800429084953981,
+      "learning_rate": 3.781985141089822e-07,
+      "loss": 1.0059,
+      "step": 19880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8974559742641277,
+      "learning_rate": 3.774761707891955e-07,
+      "loss": 1.04,
+      "step": 19881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.99561711720413,
+      "learning_rate": 3.7675451270421937e-07,
+      "loss": 1.5703,
+      "step": 19882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.100966585324804,
+      "learning_rate": 3.760335398741405e-07,
+      "loss": 1.0566,
+      "step": 19883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.506826727037572,
+      "learning_rate": 3.7531325231902615e-07,
+      "loss": 1.9102,
+      "step": 19884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.30358131145656,
+      "learning_rate": 3.745936500589187e-07,
+      "loss": 2.4365,
+      "step": 19885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5155451108544513,
+      "learning_rate": 3.738747331138437e-07,
+      "loss": 1.1348,
+      "step": 19886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.790446927132692,
+      "learning_rate": 3.731565015038102e-07,
+      "loss": 0.7695,
+      "step": 19887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.337920301053038,
+      "learning_rate": 3.72438955248805e-07,
+      "loss": 1.2754,
+      "step": 19888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.457134105185197,
+      "learning_rate": 3.71722094368801e-07,
+      "loss": 0.9248,
+      "step": 19889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4333067796730203,
+      "learning_rate": 3.7100591888374336e-07,
+      "loss": 0.73,
+      "step": 19890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.677447481041021,
+      "learning_rate": 3.7029042881356337e-07,
+      "loss": 0.6963,
+      "step": 19891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6009045934606423,
+      "learning_rate": 3.695756241781728e-07,
+      "loss": 0.6782,
+      "step": 19892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.419213945027673,
+      "learning_rate": 3.68861504997467e-07,
+      "loss": 1.6035,
+      "step": 19893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9996552170157926,
+      "learning_rate": 3.681480712913188e-07,
+      "loss": 1.3203,
+      "step": 19894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.206597577679851,
+      "learning_rate": 3.6743532307958186e-07,
+      "loss": 0.947,
+      "step": 19895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6943081518705925,
+      "learning_rate": 3.667232603820903e-07,
+      "loss": 1.4043,
+      "step": 19896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.9351565288255586,
+      "learning_rate": 3.660118832186643e-07,
+      "loss": 1.8242,
+      "step": 19897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2984736063057754,
+      "learning_rate": 3.653011916090993e-07,
+      "loss": 1.5449,
+      "step": 19898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.718003181399518,
+      "learning_rate": 3.645911855731737e-07,
+      "loss": 1.4316,
+      "step": 19899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.337879101207762,
+      "learning_rate": 3.6388186513064413e-07,
+      "loss": 1.5215,
+      "step": 19900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.074949991515369,
+      "learning_rate": 3.6317323030125304e-07,
+      "loss": 1.1846,
+      "step": 19901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.93944199972813,
+      "learning_rate": 3.624652811047236e-07,
+      "loss": 0.4526,
+      "step": 19902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.663475352269744,
+      "learning_rate": 3.617580175607538e-07,
+      "loss": 1.0625,
+      "step": 19903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.134328118807547,
+      "learning_rate": 3.61051439689028e-07,
+      "loss": 1.6309,
+      "step": 19904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.473113248275126,
+      "learning_rate": 3.60345547509211e-07,
+      "loss": 0.8398,
+      "step": 19905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.64383407711591,
+      "learning_rate": 3.5964034104094544e-07,
+      "loss": 0.604,
+      "step": 19906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.176443440945742,
+      "learning_rate": 3.589358203038601e-07,
+      "loss": 0.6021,
+      "step": 19907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5307368005508017,
+      "learning_rate": 3.5823198531755863e-07,
+      "loss": 1.0391,
+      "step": 19908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.844905276477303,
+      "learning_rate": 3.5752883610162824e-07,
+      "loss": 0.7263,
+      "step": 19909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.538220450817029,
+      "learning_rate": 3.568263726756366e-07,
+      "loss": 0.9141,
+      "step": 19910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.000169590289123,
+      "learning_rate": 3.561245950591402e-07,
+      "loss": 1.1504,
+      "step": 19911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.3575158414062,
+      "learning_rate": 3.5542350327165966e-07,
+      "loss": 1.6289,
+      "step": 19912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.216727224532701,
+      "learning_rate": 3.5472309733270983e-07,
+      "loss": 0.9951,
+      "step": 19913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.581937460371666,
+      "learning_rate": 3.5402337726178624e-07,
+      "loss": 2.1992,
+      "step": 19914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6818701749325733,
+      "learning_rate": 3.533243430783567e-07,
+      "loss": 1.3027,
+      "step": 19915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8538387880619798,
+      "learning_rate": 3.5262599480187786e-07,
+      "loss": 1.4785,
+      "step": 19916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.086042262363361,
+      "learning_rate": 3.519283324517841e-07,
+      "loss": 1.6523,
+      "step": 19917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8273930735316304,
+      "learning_rate": 3.512313560474906e-07,
+      "loss": 1.1689,
+      "step": 19918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0594545479260145,
+      "learning_rate": 3.5053506560839e-07,
+      "loss": 1.042,
+      "step": 19919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.395761096252329,
+      "learning_rate": 3.498394611538669e-07,
+      "loss": 1.1123,
+      "step": 19920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3136359852666013,
+      "learning_rate": 3.4914454270327804e-07,
+      "loss": 1.0176,
+      "step": 19921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.069494812912119,
+      "learning_rate": 3.4845031027595787e-07,
+      "loss": 0.4723,
+      "step": 19922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.563087768135443,
+      "learning_rate": 3.477567638912327e-07,
+      "loss": 1.125,
+      "step": 19923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5647353805198336,
+      "learning_rate": 3.470639035684009e-07,
+      "loss": 0.6562,
+      "step": 19924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.949495529064013,
+      "learning_rate": 3.4637172932674155e-07,
+      "loss": 0.8516,
+      "step": 19925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.933939622586732,
+      "learning_rate": 3.4568024118552254e-07,
+      "loss": 0.8955,
+      "step": 19926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1149287826415697,
+      "learning_rate": 3.4498943916398685e-07,
+      "loss": 0.4218,
+      "step": 19927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.856608902901008,
+      "learning_rate": 3.442993232813552e-07,
+      "loss": 0.4253,
+      "step": 19928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.920603753802024,
+      "learning_rate": 3.4360989355683716e-07,
+      "loss": 1.7051,
+      "step": 19929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3698105584254425,
+      "learning_rate": 3.429211500096174e-07,
+      "loss": 1.6504,
+      "step": 19930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.312102654670288,
+      "learning_rate": 3.4223309265886406e-07,
+      "loss": 1.584,
+      "step": 19931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.908999875118165,
+      "learning_rate": 3.4154572152372833e-07,
+      "loss": 0.725,
+      "step": 19932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7488460141941928,
+      "learning_rate": 3.408590366233366e-07,
+      "loss": 0.6396,
+      "step": 19933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.063128659865538,
+      "learning_rate": 3.401730379767959e-07,
+      "loss": 1.063,
+      "step": 19934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.860060781484583,
+      "learning_rate": 3.3948772560320195e-07,
+      "loss": 0.8936,
+      "step": 19935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.727022118748348,
+      "learning_rate": 3.388030995216257e-07,
+      "loss": 0.8682,
+      "step": 19936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.079481578876485,
+      "learning_rate": 3.381191597511185e-07,
+      "loss": 1.0298,
+      "step": 19937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.64468162480977,
+      "learning_rate": 3.3743590631071795e-07,
+      "loss": 0.8975,
+      "step": 19938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.785867161240299,
+      "learning_rate": 3.3675333921943107e-07,
+      "loss": 1.3311,
+      "step": 19939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.051614995140987,
+      "learning_rate": 3.360714584962621e-07,
+      "loss": 1.3125,
+      "step": 19940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.658748663223569,
+      "learning_rate": 3.353902641601847e-07,
+      "loss": 1.0391,
+      "step": 19941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.209969352406096,
+      "learning_rate": 3.347097562301532e-07,
+      "loss": 1.1357,
+      "step": 19942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.142671403861112,
+      "learning_rate": 3.34029934725108e-07,
+      "loss": 1.0498,
+      "step": 19943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2926447312855958,
+      "learning_rate": 3.333507996639701e-07,
+      "loss": 0.5867,
+      "step": 19944
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6122446493780782,
+      "learning_rate": 3.3267235106563545e-07,
+      "loss": 1.3691,
+      "step": 19945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0963675474878207,
+      "learning_rate": 3.3199458894898614e-07,
+      "loss": 1.0444,
+      "step": 19946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.273146043690023,
+      "learning_rate": 3.313175133328877e-07,
+      "loss": 0.6394,
+      "step": 19947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8512427460337102,
+      "learning_rate": 3.306411242361779e-07,
+      "loss": 0.8369,
+      "step": 19948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1755440073208936,
+      "learning_rate": 3.2996542167768596e-07,
+      "loss": 0.917,
+      "step": 19949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4635670749430267,
+      "learning_rate": 3.2929040567621085e-07,
+      "loss": 0.7031,
+      "step": 19950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.01693217453925,
+      "learning_rate": 3.2861607625054027e-07,
+      "loss": 0.9648,
+      "step": 19951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.272347100279077,
+      "learning_rate": 3.279424334194425e-07,
+      "loss": 1.1748,
+      "step": 19952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5032196293730236,
+      "learning_rate": 3.27269477201661e-07,
+      "loss": 0.6724,
+      "step": 19953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2031088754050328,
+      "learning_rate": 3.2659720761592795e-07,
+      "loss": 1.2031,
+      "step": 19954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.39458392086877,
+      "learning_rate": 3.25925624680945e-07,
+      "loss": 0.7046,
+      "step": 19955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5309188822565964,
+      "learning_rate": 3.2525472841541114e-07,
+      "loss": 1.0078,
+      "step": 19956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.212593024932948,
+      "learning_rate": 3.24584518837992e-07,
+      "loss": 0.6194,
+      "step": 19957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.888960777582615,
+      "learning_rate": 3.2391499596734207e-07,
+      "loss": 1.4121,
+      "step": 19958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.302132434810388,
+      "learning_rate": 3.232461598220909e-07,
+      "loss": 1.2236,
+      "step": 19959
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.791012606258458,
+      "learning_rate": 3.2257801042085144e-07,
+      "loss": 0.7672,
+      "step": 19960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5392640958129804,
+      "learning_rate": 3.219105477822226e-07,
+      "loss": 0.8701,
+      "step": 19961
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1913497835931977,
+      "learning_rate": 3.2124377192477575e-07,
+      "loss": 1.1523,
+      "step": 19962
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5700869939604494,
+      "learning_rate": 3.205776828670681e-07,
+      "loss": 1.1377,
+      "step": 19963
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2694450558474855,
+      "learning_rate": 3.19912280627635e-07,
+      "loss": 1.2344,
+      "step": 19964
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1900581660450027,
+      "learning_rate": 3.192475652249949e-07,
+      "loss": 1.1191,
+      "step": 19965
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3114231798599847,
+      "learning_rate": 3.185835366776496e-07,
+      "loss": 1.1836,
+      "step": 19966
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4936758159416406,
+      "learning_rate": 3.179201950040789e-07,
+      "loss": 0.5195,
+      "step": 19967
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.546420030359952,
+      "learning_rate": 3.1725754022273737e-07,
+      "loss": 0.6201,
+      "step": 19968
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5064213301775293,
+      "learning_rate": 3.1659557235207136e-07,
+      "loss": 1.1074,
+      "step": 19969
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.410454146243503,
+      "learning_rate": 3.1593429141050515e-07,
+      "loss": 1.3086,
+      "step": 19970
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.73994705448496,
+      "learning_rate": 3.1527369741643497e-07,
+      "loss": 1.1836,
+      "step": 19971
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.352547879517257,
+      "learning_rate": 3.14613790388249e-07,
+      "loss": 1.4199,
+      "step": 19972
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2621325705591766,
+      "learning_rate": 3.139545703443131e-07,
+      "loss": 0.759,
+      "step": 19973
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7057477982839977,
+      "learning_rate": 3.132960373029709e-07,
+      "loss": 1.1748,
+      "step": 19974
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.533901410327872,
+      "learning_rate": 3.1263819128254936e-07,
+      "loss": 0.2141,
+      "step": 19975
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.09272345564953,
+      "learning_rate": 3.1198103230135887e-07,
+      "loss": 0.959,
+      "step": 19976
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.728625608153928,
+      "learning_rate": 3.113245603776821e-07,
+      "loss": 1.0117,
+      "step": 19977
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.804637165110753,
+      "learning_rate": 3.10668775529796e-07,
+      "loss": 1.332,
+      "step": 19978
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.380046702656007,
+      "learning_rate": 3.100136777759444e-07,
+      "loss": 0.6677,
+      "step": 19979
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.893488524017644,
+      "learning_rate": 3.0935926713436256e-07,
+      "loss": 1.0615,
+      "step": 19980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5879287659335577,
+      "learning_rate": 3.087055436232583e-07,
+      "loss": 1.229,
+      "step": 19981
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5574538625543646,
+      "learning_rate": 3.080525072608309e-07,
+      "loss": 0.832,
+      "step": 19982
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.264305468729339,
+      "learning_rate": 3.0740015806524637e-07,
+      "loss": 1.1445,
+      "step": 19983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.271507256508248,
+      "learning_rate": 3.0674849605466514e-07,
+      "loss": 1.2217,
+      "step": 19984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.039722853492827,
+      "learning_rate": 3.0609752124721723e-07,
+      "loss": 1.0762,
+      "step": 19985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.344290720265233,
+      "learning_rate": 3.05447233661027e-07,
+      "loss": 0.9717,
+      "step": 19986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.926010061772139,
+      "learning_rate": 3.0479763331418556e-07,
+      "loss": 0.8994,
+      "step": 19987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0451768104070123,
+      "learning_rate": 3.041487202247728e-07,
+      "loss": 0.9414,
+      "step": 19988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3988351949019362,
+      "learning_rate": 3.0350049441084383e-07,
+      "loss": 0.2217,
+      "step": 19989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6686995526086306,
+      "learning_rate": 3.0285295589044806e-07,
+      "loss": 0.96,
+      "step": 19990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8812038842746452,
+      "learning_rate": 3.0220610468159604e-07,
+      "loss": 1.6777,
+      "step": 19991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.599210024463016,
+      "learning_rate": 3.0155994080229563e-07,
+      "loss": 0.9463,
+      "step": 19992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.136324049289573,
+      "learning_rate": 3.009144642705269e-07,
+      "loss": 0.7744,
+      "step": 19993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9445313873229662,
+      "learning_rate": 3.002696751042533e-07,
+      "loss": 0.877,
+      "step": 19994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1030355630027158,
+      "learning_rate": 2.996255733214187e-07,
+      "loss": 1.2363,
+      "step": 19995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5724382145038915,
+      "learning_rate": 2.9898215893995054e-07,
+      "loss": 1.2861,
+      "step": 19996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.164750160818126,
+      "learning_rate": 2.983394319777538e-07,
+      "loss": 1.2871,
+      "step": 19997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0231044892929204,
+      "learning_rate": 2.9769739245271157e-07,
+      "loss": 0.6531,
+      "step": 19998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7514604678453445,
+      "learning_rate": 2.9705604038269554e-07,
+      "loss": 1.2129,
+      "step": 19999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9187650251266228,
+      "learning_rate": 2.9641537578555536e-07,
+      "loss": 1.0547,
+      "step": 20000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.631347248518362,
+      "learning_rate": 2.957753986791184e-07,
+      "loss": 0.8105,
+      "step": 20001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.421303269366998,
+      "learning_rate": 2.9513610908118996e-07,
+      "loss": 0.6115,
+      "step": 20002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3417101311863893,
+      "learning_rate": 2.9449750700956957e-07,
+      "loss": 1.207,
+      "step": 20003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.433044186408733,
+      "learning_rate": 2.9385959248202645e-07,
+      "loss": 0.9678,
+      "step": 20004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.633285523566599,
+      "learning_rate": 2.932223655163158e-07,
+      "loss": 1.186,
+      "step": 20005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.185990602994861,
+      "learning_rate": 2.925858261301623e-07,
+      "loss": 1.2559,
+      "step": 20006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9119528326050212,
+      "learning_rate": 2.9194997434129067e-07,
+      "loss": 0.5446,
+      "step": 20007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.037522751121077,
+      "learning_rate": 2.9131481016739516e-07,
+      "loss": 0.9868,
+      "step": 20008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5611516174245033,
+      "learning_rate": 2.906803336261449e-07,
+      "loss": 0.9033,
+      "step": 20009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.819605857684366,
+      "learning_rate": 2.900465447352035e-07,
+      "loss": 1.1445,
+      "step": 20010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.05093909870319,
+      "learning_rate": 2.89413443512207e-07,
+      "loss": 1.293,
+      "step": 20011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3034531213367564,
+      "learning_rate": 2.887810299747745e-07,
+      "loss": 1.0288,
+      "step": 20012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.309448806760207,
+      "learning_rate": 2.881493041405059e-07,
+      "loss": 0.8279,
+      "step": 20013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4475630376255144,
+      "learning_rate": 2.875182660269843e-07,
+      "loss": 0.8213,
+      "step": 20014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.313820436169095,
+      "learning_rate": 2.8688791565176507e-07,
+      "loss": 0.5059,
+      "step": 20015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0692287405715923,
+      "learning_rate": 2.8625825303239815e-07,
+      "loss": 1.2969,
+      "step": 20016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.617796414268848,
+      "learning_rate": 2.8562927818640286e-07,
+      "loss": 1.1426,
+      "step": 20017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.45865755892771,
+      "learning_rate": 2.8500099113128176e-07,
+      "loss": 1.1337,
+      "step": 20018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.04325406651458,
+      "learning_rate": 2.8437339188452095e-07,
+      "loss": 1.668,
+      "step": 20019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.350668051994008,
+      "learning_rate": 2.83746480463587e-07,
+      "loss": 1.4863,
+      "step": 20020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.6037456517831865,
+      "learning_rate": 2.8312025688592704e-07,
+      "loss": 1.1221,
+      "step": 20021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.079805022174192,
+      "learning_rate": 2.824947211689688e-07,
+      "loss": 1.3984,
+      "step": 20022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.676714256959346,
+      "learning_rate": 2.8186987333011784e-07,
+      "loss": 1.1172,
+      "step": 20023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.223287219423295,
+      "learning_rate": 2.8124571338676855e-07,
+      "loss": 1.0938,
+      "step": 20024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7967921118460217,
+      "learning_rate": 2.806222413562876e-07,
+      "loss": 1.418,
+      "step": 20025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.958313736222379,
+      "learning_rate": 2.7999945725602493e-07,
+      "loss": 0.9756,
+      "step": 20026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.376198304782726,
+      "learning_rate": 2.7937736110331405e-07,
+      "loss": 1.5742,
+      "step": 20027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0152144869145894,
+      "learning_rate": 2.7875595291546873e-07,
+      "loss": 1.1934,
+      "step": 20028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.001436790312513,
+      "learning_rate": 2.7813523270978083e-07,
+      "loss": 1.0381,
+      "step": 20029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7251150331080645,
+      "learning_rate": 2.775152005035253e-07,
+      "loss": 1.2148,
+      "step": 20030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.416507766072456,
+      "learning_rate": 2.768958563139551e-07,
+      "loss": 1.0605,
+      "step": 20031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.530447279252218,
+      "learning_rate": 2.762772001583092e-07,
+      "loss": 0.7926,
+      "step": 20032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9924997122464039,
+      "learning_rate": 2.7565923205380707e-07,
+      "loss": 0.4674,
+      "step": 20033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.908707966909405,
+      "learning_rate": 2.750419520176406e-07,
+      "loss": 1.168,
+      "step": 20034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.30512070544577,
+      "learning_rate": 2.7442536006699047e-07,
+      "loss": 0.9248,
+      "step": 20035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.204970107968013,
+      "learning_rate": 2.7380945621901797e-07,
+      "loss": 1.1074,
+      "step": 20036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8946754007458069,
+      "learning_rate": 2.7319424049086216e-07,
+      "loss": 1.229,
+      "step": 20037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.407914712903661,
+      "learning_rate": 2.725797128996427e-07,
+      "loss": 1.5605,
+      "step": 20038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9298574169475513,
+      "learning_rate": 2.7196587346246525e-07,
+      "loss": 1.5703,
+      "step": 20039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.914488896732119,
+      "learning_rate": 2.713527221964107e-07,
+      "loss": 0.998,
+      "step": 20040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.057228866455619,
+      "learning_rate": 2.707402591185404e-07,
+      "loss": 1.3633,
+      "step": 20041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4027998328498286,
+      "learning_rate": 2.701284842459045e-07,
+      "loss": 0.8945,
+      "step": 20042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.61307703123206,
+      "learning_rate": 2.6951739759552285e-07,
+      "loss": 0.6582,
+      "step": 20043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.344972927205641,
+      "learning_rate": 2.6890699918440676e-07,
+      "loss": 0.8516,
+      "step": 20044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8374670921025869,
+      "learning_rate": 2.682972890295399e-07,
+      "loss": 0.9258,
+      "step": 20045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.741445769319569,
+      "learning_rate": 2.676882671478892e-07,
+      "loss": 1.6465,
+      "step": 20046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.764332787305495,
+      "learning_rate": 2.670799335564078e-07,
+      "loss": 1.0088,
+      "step": 20047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.562121022827947,
+      "learning_rate": 2.6647228827202385e-07,
+      "loss": 1.084,
+      "step": 20048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.769613896877586,
+      "learning_rate": 2.6586533131164326e-07,
+      "loss": 1.293,
+      "step": 20049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7061885166011685,
+      "learning_rate": 2.6525906269216363e-07,
+      "loss": 1.0527,
+      "step": 20050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.606125919039322,
+      "learning_rate": 2.6465348243045474e-07,
+      "loss": 1.1584,
+      "step": 20051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.024295182519918,
+      "learning_rate": 2.640485905433726e-07,
+      "loss": 0.8994,
+      "step": 20052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.822829109308057,
+      "learning_rate": 2.6344438704774266e-07,
+      "loss": 1.1143,
+      "step": 20053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.813849639192046,
+      "learning_rate": 2.6284087196038753e-07,
+      "loss": 1.7217,
+      "step": 20054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.798935115496913,
+      "learning_rate": 2.6223804529810213e-07,
+      "loss": 1.5684,
+      "step": 20055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8646203795840766,
+      "learning_rate": 2.6163590707766194e-07,
+      "loss": 1.0498,
+      "step": 20056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5626499783425407,
+      "learning_rate": 2.6103445731582023e-07,
+      "loss": 1.3457,
+      "step": 20057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.761947600771593,
+      "learning_rate": 2.604336960293163e-07,
+      "loss": 0.7603,
+      "step": 20058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.634275488371103,
+      "learning_rate": 2.598336232348758e-07,
+      "loss": 1.2256,
+      "step": 20059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5891311148355487,
+      "learning_rate": 2.5923423894919364e-07,
+      "loss": 1.5449,
+      "step": 20060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.196221674203578,
+      "learning_rate": 2.5863554318894814e-07,
+      "loss": 1.3027,
+      "step": 20061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.062609861795759,
+      "learning_rate": 2.5803753597080374e-07,
+      "loss": 0.6426,
+      "step": 20062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.707976205216959,
+      "learning_rate": 2.5744021731140266e-07,
+      "loss": 0.7852,
+      "step": 20063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.827885763458634,
+      "learning_rate": 2.5684358722737054e-07,
+      "loss": 0.9734,
+      "step": 20064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.911598124870098,
+      "learning_rate": 2.562476457353025e-07,
+      "loss": 0.7881,
+      "step": 20065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.797567808338966,
+      "learning_rate": 2.556523928517934e-07,
+      "loss": 0.8584,
+      "step": 20066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6202494004624404,
+      "learning_rate": 2.5505782859340245e-07,
+      "loss": 1.0898,
+      "step": 20067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.239929609973989,
+      "learning_rate": 2.544639529766829e-07,
+      "loss": 0.4924,
+      "step": 20068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.076959414870275,
+      "learning_rate": 2.5387076601815494e-07,
+      "loss": 1.0513,
+      "step": 20069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6161358496058131,
+      "learning_rate": 2.532782677343276e-07,
+      "loss": 0.7301,
+      "step": 20070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6657282747787394,
+      "learning_rate": 2.5268645814169324e-07,
+      "loss": 1.5215,
+      "step": 20071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.238653577934082,
+      "learning_rate": 2.52095337256722e-07,
+      "loss": 1.832,
+      "step": 20072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.547062591945875,
+      "learning_rate": 2.51504905095859e-07,
+      "loss": 1.6113,
+      "step": 20073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.15879436720962,
+      "learning_rate": 2.5091516167554117e-07,
+      "loss": 1.2559,
+      "step": 20074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0088770184015887,
+      "learning_rate": 2.503261070121804e-07,
+      "loss": 0.4995,
+      "step": 20075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.152505749932414,
+      "learning_rate": 2.497377411221663e-07,
+      "loss": 1.168,
+      "step": 20076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6927280679188155,
+      "learning_rate": 2.491500640218747e-07,
+      "loss": 0.7356,
+      "step": 20077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.441699130418568,
+      "learning_rate": 2.485630757276619e-07,
+      "loss": 1.6855,
+      "step": 20078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.621032648486651,
+      "learning_rate": 2.479767762558621e-07,
+      "loss": 1.0332,
+      "step": 20079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.550331438253916,
+      "learning_rate": 2.4739116562279287e-07,
+      "loss": 0.9026,
+      "step": 20080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.026660998216222,
+      "learning_rate": 2.4680624384475215e-07,
+      "loss": 1.4512,
+      "step": 20081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.778710481034044,
+      "learning_rate": 2.462220109380131e-07,
+      "loss": 0.5605,
+      "step": 20082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.86993059865918,
+      "learning_rate": 2.456384669188405e-07,
+      "loss": 0.8652,
+      "step": 20083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.554848931463276,
+      "learning_rate": 2.450556118034714e-07,
+      "loss": 0.7084,
+      "step": 20084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.196337652003875,
+      "learning_rate": 2.44473445608126e-07,
+      "loss": 1.3281,
+      "step": 20085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.576276009927384,
+      "learning_rate": 2.438919683490054e-07,
+      "loss": 1.5176,
+      "step": 20086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.245979316245959,
+      "learning_rate": 2.433111800422938e-07,
+      "loss": 0.6357,
+      "step": 20087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4192013681429536,
+      "learning_rate": 2.4273108070415607e-07,
+      "loss": 0.959,
+      "step": 20088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5315885906054603,
+      "learning_rate": 2.4215167035073205e-07,
+      "loss": 1.1279,
+      "step": 20089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3023426606040114,
+      "learning_rate": 2.415729489981505e-07,
+      "loss": 1.1414,
+      "step": 20090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.78548946543905,
+      "learning_rate": 2.409949166625125e-07,
+      "loss": 1.2539,
+      "step": 20091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0003579534638503,
+      "learning_rate": 2.404175733599051e-07,
+      "loss": 0.8984,
+      "step": 20092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.934672354195107,
+      "learning_rate": 2.3984091910639883e-07,
+      "loss": 1.046,
+      "step": 20093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.83461324916449,
+      "learning_rate": 2.3926495391804195e-07,
+      "loss": 0.9131,
+      "step": 20094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.699573776580253,
+      "learning_rate": 2.3868967781085773e-07,
+      "loss": 0.5815,
+      "step": 20095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.7812513962288605,
+      "learning_rate": 2.3811509080085837e-07,
+      "loss": 0.7666,
+      "step": 20096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.983421688933852,
+      "learning_rate": 2.3754119290403943e-07,
+      "loss": 1.1562,
+      "step": 20097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.059754133273817,
+      "learning_rate": 2.369679841363659e-07,
+      "loss": 1.459,
+      "step": 20098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3279546416474477,
+      "learning_rate": 2.3639546451379447e-07,
+      "loss": 0.9863,
+      "step": 20099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.878355192120923,
+      "learning_rate": 2.3582363405225405e-07,
+      "loss": 0.8311,
+      "step": 20100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5307691128505243,
+      "learning_rate": 2.352524927676597e-07,
+      "loss": 0.8477,
+      "step": 20101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.112253208410935,
+      "learning_rate": 2.3468204067590982e-07,
+      "loss": 0.7803,
+      "step": 20102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.239747424361252,
+      "learning_rate": 2.3411227779287226e-07,
+      "loss": 1.1826,
+      "step": 20103
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4738305845237853,
+      "learning_rate": 2.335432041344121e-07,
+      "loss": 1.2383,
+      "step": 20104
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4289278762556448,
+      "learning_rate": 2.3297481971635838e-07,
+      "loss": 1.3047,
+      "step": 20105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.0748280312853105,
+      "learning_rate": 2.3240712455453727e-07,
+      "loss": 0.8108,
+      "step": 20106
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3339024216915587,
+      "learning_rate": 2.3184011866474176e-07,
+      "loss": 1.1934,
+      "step": 20107
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8457735009743232,
+      "learning_rate": 2.3127380206274807e-07,
+      "loss": 0.8633,
+      "step": 20108
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.392253944077192,
+      "learning_rate": 2.3070817476432693e-07,
+      "loss": 1.2725,
+      "step": 20109
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.184416859761871,
+      "learning_rate": 2.3014323678521022e-07,
+      "loss": 1.4609,
+      "step": 20110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.786948121183758,
+      "learning_rate": 2.29578988141127e-07,
+      "loss": 1.2227,
+      "step": 20111
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.3938154799030364,
+      "learning_rate": 2.2901542884777305e-07,
+      "loss": 0.6943,
+      "step": 20112
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.324163580739482,
+      "learning_rate": 2.2845255892083583e-07,
+      "loss": 1.2246,
+      "step": 20113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.707638715041918,
+      "learning_rate": 2.2789037837597781e-07,
+      "loss": 1.1602,
+      "step": 20114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4672033559472086,
+      "learning_rate": 2.2732888722885038e-07,
+      "loss": 0.7915,
+      "step": 20115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.116921320284762,
+      "learning_rate": 2.2676808549507157e-07,
+      "loss": 1.3008,
+      "step": 20116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.24035945970514,
+      "learning_rate": 2.262079731902539e-07,
+      "loss": 1.2637,
+      "step": 20117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.572926144021655,
+      "learning_rate": 2.256485503299821e-07,
+      "loss": 1.2061,
+      "step": 20118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.388205221702716,
+      "learning_rate": 2.2508981692982434e-07,
+      "loss": 1.5137,
+      "step": 20119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2575136402685443,
+      "learning_rate": 2.24531773005332e-07,
+      "loss": 1.043,
+      "step": 20120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.920343921177727,
+      "learning_rate": 2.2397441857203438e-07,
+      "loss": 0.8994,
+      "step": 20121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7171302922493994,
+      "learning_rate": 2.2341775364544403e-07,
+      "loss": 0.9473,
+      "step": 20122
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5657387711718687,
+      "learning_rate": 2.228617782410486e-07,
+      "loss": 0.964,
+      "step": 20123
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3583624479655625,
+      "learning_rate": 2.2230649237432177e-07,
+      "loss": 0.8599,
+      "step": 20124
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6855361327910154,
+      "learning_rate": 2.2175189606072067e-07,
+      "loss": 1.1953,
+      "step": 20125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0387740241716252,
+      "learning_rate": 2.2119798931567738e-07,
+      "loss": 1.3496,
+      "step": 20126
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9294863093844277,
+      "learning_rate": 2.2064477215460733e-07,
+      "loss": 1.2402,
+      "step": 20127
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.49914785638235,
+      "learning_rate": 2.200922445929038e-07,
+      "loss": 1.1621,
+      "step": 20128
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5210578139633637,
+      "learning_rate": 2.1954040664594334e-07,
+      "loss": 0.4478,
+      "step": 20129
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.842385573642731,
+      "learning_rate": 2.1898925832908867e-07,
+      "loss": 1.3945,
+      "step": 20130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.080791761142854,
+      "learning_rate": 2.184387996576748e-07,
+      "loss": 2.0117,
+      "step": 20131
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.610027303830182,
+      "learning_rate": 2.1788903064701716e-07,
+      "loss": 1.0508,
+      "step": 20132
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.641238769889368,
+      "learning_rate": 2.1733995131242301e-07,
+      "loss": 1.4502,
+      "step": 20133
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.655554646749443,
+      "learning_rate": 2.1679156166916624e-07,
+      "loss": 1.1211,
+      "step": 20134
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1959762070477975,
+      "learning_rate": 2.1624386173251244e-07,
+      "loss": 1.1523,
+      "step": 20135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.328112678325929,
+      "learning_rate": 2.1569685151770213e-07,
+      "loss": 1.0088,
+      "step": 20136
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.258237821001777,
+      "learning_rate": 2.151505310399593e-07,
+      "loss": 0.8936,
+      "step": 20137
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7611227782565286,
+      "learning_rate": 2.1460490031448567e-07,
+      "loss": 1.0977,
+      "step": 20138
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.750510815031049,
+      "learning_rate": 2.1405995935646906e-07,
+      "loss": 1.1514,
+      "step": 20139
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6273139471071159,
+      "learning_rate": 2.1351570818107237e-07,
+      "loss": 0.9843,
+      "step": 20140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4833560018090823,
+      "learning_rate": 2.129721468034418e-07,
+      "loss": 0.9043,
+      "step": 20141
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.766300341172464,
+      "learning_rate": 2.1242927523870693e-07,
+      "loss": 0.7822,
+      "step": 20142
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.326150514418713,
+      "learning_rate": 2.1188709350197511e-07,
+      "loss": 0.6934,
+      "step": 20143
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4143697302412943,
+      "learning_rate": 2.1134560160833427e-07,
+      "loss": 1.0166,
+      "step": 20144
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.037124355769461,
+      "learning_rate": 2.1080479957285293e-07,
+      "loss": 0.8267,
+      "step": 20145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.963200570781036,
+      "learning_rate": 2.1026468741058015e-07,
+      "loss": 1.3799,
+      "step": 20146
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8114819697521547,
+      "learning_rate": 2.0972526513655388e-07,
+      "loss": 0.6354,
+      "step": 20147
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.294990264917073,
+      "learning_rate": 2.0918653276577883e-07,
+      "loss": 1.0029,
+      "step": 20148
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1275496053683973,
+      "learning_rate": 2.086484903132513e-07,
+      "loss": 0.376,
+      "step": 20149
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.025640114137959,
+      "learning_rate": 2.0811113779394264e-07,
+      "loss": 0.8447,
+      "step": 20150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3619743327491247,
+      "learning_rate": 2.075744752228076e-07,
+      "loss": 0.8291,
+      "step": 20151
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5273821443120648,
+      "learning_rate": 2.070385026147814e-07,
+      "loss": 1.5742,
+      "step": 20152
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4479983268806875,
+      "learning_rate": 2.0650321998478272e-07,
+      "loss": 0.916,
+      "step": 20153
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7564088529106665,
+      "learning_rate": 2.0596862734770518e-07,
+      "loss": 1.416,
+      "step": 20154
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2063560007282037,
+      "learning_rate": 2.0543472471842573e-07,
+      "loss": 1.2695,
+      "step": 20155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7738836225150694,
+      "learning_rate": 2.0490151211180752e-07,
+      "loss": 0.8838,
+      "step": 20156
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.725090175851474,
+      "learning_rate": 2.0436898954268314e-07,
+      "loss": 1.1299,
+      "step": 20157
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.204296083370802,
+      "learning_rate": 2.038371570258768e-07,
+      "loss": 1.0381,
+      "step": 20158
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.497313009141025,
+      "learning_rate": 2.0330601457618781e-07,
+      "loss": 1.2656,
+      "step": 20159
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.896389899387236,
+      "learning_rate": 2.02775562208396e-07,
+      "loss": 1.4199,
+      "step": 20160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3919682105482103,
+      "learning_rate": 2.0224579993726732e-07,
+      "loss": 1.2422,
+      "step": 20161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.395803191506782,
+      "learning_rate": 2.017167277775428e-07,
+      "loss": 0.998,
+      "step": 20162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.433812718298696,
+      "learning_rate": 2.0118834574394395e-07,
+      "loss": 1.4043,
+      "step": 20163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2945655059075736,
+      "learning_rate": 2.0066065385118128e-07,
+      "loss": 1.248,
+      "step": 20164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2094003031143665,
+      "learning_rate": 2.0013365211393463e-07,
+      "loss": 0.7515,
+      "step": 20165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7994062361364285,
+      "learning_rate": 1.996073405468729e-07,
+      "loss": 1.0713,
+      "step": 20166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7566048097426705,
+      "learning_rate": 1.9908171916464269e-07,
+      "loss": 0.9595,
+      "step": 20167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7439504319932757,
+      "learning_rate": 1.9855678798187116e-07,
+      "loss": 0.7197,
+      "step": 20168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0185778376151933,
+      "learning_rate": 1.980325470131661e-07,
+      "loss": 0.5051,
+      "step": 20169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.669777491821224,
+      "learning_rate": 1.9750899627312136e-07,
+      "loss": 0.623,
+      "step": 20170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.389671183141513,
+      "learning_rate": 1.969861357763031e-07,
+      "loss": 1.8848,
+      "step": 20171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7654835293034,
+      "learning_rate": 1.9646396553726353e-07,
+      "loss": 0.8799,
+      "step": 20172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4472924596158103,
+      "learning_rate": 1.9594248557053551e-07,
+      "loss": 0.834,
+      "step": 20173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8911353408237326,
+      "learning_rate": 1.9542169589062965e-07,
+      "loss": 1.0151,
+      "step": 20174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.1073341143272675,
+      "learning_rate": 1.9490159651203988e-07,
+      "loss": 1.1963,
+      "step": 20175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.752741775262382,
+      "learning_rate": 1.9438218744923797e-07,
+      "loss": 1.5762,
+      "step": 20176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.68368177420013,
+      "learning_rate": 1.938634687166846e-07,
+      "loss": 0.9121,
+      "step": 20177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9358265941216286,
+      "learning_rate": 1.9334544032880985e-07,
+      "loss": 0.8997,
+      "step": 20178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.683525207965266,
+      "learning_rate": 1.9282810230003557e-07,
+      "loss": 0.6694,
+      "step": 20179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.583474821912639,
+      "learning_rate": 1.9231145464475297e-07,
+      "loss": 0.5933,
+      "step": 20180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.670114557832595,
+      "learning_rate": 1.9179549737734503e-07,
+      "loss": 0.7793,
+      "step": 20181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1112352490257424,
+      "learning_rate": 1.912802305121697e-07,
+      "loss": 0.6431,
+      "step": 20182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.711358425963032,
+      "learning_rate": 1.907656540635655e-07,
+      "loss": 0.9194,
+      "step": 20183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.699224034983355,
+      "learning_rate": 1.9025176804585432e-07,
+      "loss": 1.2031,
+      "step": 20184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0026421975309985,
+      "learning_rate": 1.8973857247333303e-07,
+      "loss": 0.4854,
+      "step": 20185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8972311877019195,
+      "learning_rate": 1.8922606736029024e-07,
+      "loss": 1.3301,
+      "step": 20186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1351457149934285,
+      "learning_rate": 1.8871425272098676e-07,
+      "loss": 1.6309,
+      "step": 20187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.211485562281924,
+      "learning_rate": 1.8820312856966115e-07,
+      "loss": 1.041,
+      "step": 20188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.245815730925342,
+      "learning_rate": 1.8769269492054097e-07,
+      "loss": 0.7976,
+      "step": 20189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.077586170418716,
+      "learning_rate": 1.871829517878343e-07,
+      "loss": 1.1704,
+      "step": 20190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5558750763693165,
+      "learning_rate": 1.8667389918572698e-07,
+      "loss": 0.7041,
+      "step": 20191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.300425113689025,
+      "learning_rate": 1.8616553712837993e-07,
+      "loss": 1.3379,
+      "step": 20192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.431288082835609,
+      "learning_rate": 1.856578656299457e-07,
+      "loss": 1.1484,
+      "step": 20193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9158973462910973,
+      "learning_rate": 1.8515088470455188e-07,
+      "loss": 0.7173,
+      "step": 20194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.896734847746743,
+      "learning_rate": 1.8464459436630388e-07,
+      "loss": 0.7842,
+      "step": 20195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.578258010603429,
+      "learning_rate": 1.8413899462929874e-07,
+      "loss": 1.4453,
+      "step": 20196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6007543886043134,
+      "learning_rate": 1.8363408550759742e-07,
+      "loss": 1.1211,
+      "step": 20197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3210433682289753,
+      "learning_rate": 1.831298670152609e-07,
+      "loss": 0.7153,
+      "step": 20198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1627489919234955,
+      "learning_rate": 1.8262633916631687e-07,
+      "loss": 1.293,
+      "step": 20199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.543715359419261,
+      "learning_rate": 1.8212350197477635e-07,
+      "loss": 0.7407,
+      "step": 20200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.360089023495093,
+      "learning_rate": 1.8162135545463644e-07,
+      "loss": 1.793,
+      "step": 20201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1739845213870206,
+      "learning_rate": 1.8111989961987208e-07,
+      "loss": 1.3369,
+      "step": 20202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3872247337746466,
+      "learning_rate": 1.80619134484436e-07,
+      "loss": 1.0322,
+      "step": 20203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0704008263514075,
+      "learning_rate": 1.801190600622671e-07,
+      "loss": 0.9785,
+      "step": 20204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.860720171917788,
+      "learning_rate": 1.7961967636727916e-07,
+      "loss": 1.0254,
+      "step": 20205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.149827098547672,
+      "learning_rate": 1.7912098341336947e-07,
+      "loss": 0.7197,
+      "step": 20206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0620811499492038,
+      "learning_rate": 1.7862298121442134e-07,
+      "loss": 1.0215,
+      "step": 20207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3420865832692055,
+      "learning_rate": 1.7812566978429034e-07,
+      "loss": 0.748,
+      "step": 20208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.126107500696001,
+      "learning_rate": 1.7762904913681544e-07,
+      "loss": 0.7314,
+      "step": 20209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.879891899975279,
+      "learning_rate": 1.771331192858189e-07,
+      "loss": 1.4727,
+      "step": 20210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.8828707027781135,
+      "learning_rate": 1.7663788024510353e-07,
+      "loss": 1.1699,
+      "step": 20211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.745959695118775,
+      "learning_rate": 1.7614333202845001e-07,
+      "loss": 0.9551,
+      "step": 20212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.57200068735394,
+      "learning_rate": 1.7564947464962235e-07,
+      "loss": 1.1016,
+      "step": 20213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.52248360651439,
+      "learning_rate": 1.751563081223623e-07,
+      "loss": 1.6719,
+      "step": 20214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5788749342045656,
+      "learning_rate": 1.7466383246039775e-07,
+      "loss": 1.1953,
+      "step": 20215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8634115689367157,
+      "learning_rate": 1.7417204767742888e-07,
+      "loss": 1.0399,
+      "step": 20216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.533895666131297,
+      "learning_rate": 1.7368095378715033e-07,
+      "loss": 0.7061,
+      "step": 20217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5636735415391505,
+      "learning_rate": 1.7319055080322055e-07,
+      "loss": 1.0811,
+      "step": 20218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.0129288108551355,
+      "learning_rate": 1.7270083873929255e-07,
+      "loss": 1.0771,
+      "step": 20219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6876723655606516,
+      "learning_rate": 1.7221181760899152e-07,
+      "loss": 1.165,
+      "step": 20220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.058438543204603,
+      "learning_rate": 1.717234874259288e-07,
+      "loss": 0.7947,
+      "step": 20221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.387237544291442,
+      "learning_rate": 1.7123584820369355e-07,
+      "loss": 0.7432,
+      "step": 20222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.472001072940573,
+      "learning_rate": 1.707488999558582e-07,
+      "loss": 1.1182,
+      "step": 20223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.229737392056594,
+      "learning_rate": 1.7026264269597303e-07,
+      "loss": 1.1133,
+      "step": 20224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.642810075480691,
+      "learning_rate": 1.6977707643756614e-07,
+      "loss": 0.8115,
+      "step": 20225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.637391238633079,
+      "learning_rate": 1.6929220119416e-07,
+      "loss": 1.3301,
+      "step": 20226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.478448673002334,
+      "learning_rate": 1.688080169792411e-07,
+      "loss": 1.0088,
+      "step": 20227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3977735850773814,
+      "learning_rate": 1.6832452380628472e-07,
+      "loss": 0.499,
+      "step": 20228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7969196678302897,
+      "learning_rate": 1.6784172168874956e-07,
+      "loss": 0.7578,
+      "step": 20229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1817481578802953,
+      "learning_rate": 1.6735961064007212e-07,
+      "loss": 1.402,
+      "step": 20230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.56759499898775,
+      "learning_rate": 1.6687819067366383e-07,
+      "loss": 1.0527,
+      "step": 20231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9931253058456324,
+      "learning_rate": 1.6639746180292792e-07,
+      "loss": 1.0176,
+      "step": 20232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.167268073866006,
+      "learning_rate": 1.6591742404124256e-07,
+      "loss": 1.4121,
+      "step": 20233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.819732979901112,
+      "learning_rate": 1.6543807740196093e-07,
+      "loss": 1.2549,
+      "step": 20234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7525177282941333,
+      "learning_rate": 1.6495942189843073e-07,
+      "loss": 0.9297,
+      "step": 20235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7551069523428047,
+      "learning_rate": 1.6448145754396627e-07,
+      "loss": 1.3848,
+      "step": 20236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.384789419094544,
+      "learning_rate": 1.6400418435187636e-07,
+      "loss": 1.3086,
+      "step": 20237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.256065458322051,
+      "learning_rate": 1.635276023354365e-07,
+      "loss": 0.5076,
+      "step": 20238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5041918896649418,
+      "learning_rate": 1.6305171150791386e-07,
+      "loss": 0.5686,
+      "step": 20239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.243998412294242,
+      "learning_rate": 1.6257651188255064e-07,
+      "loss": 0.8477,
+      "step": 20240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3975265711484393,
+      "learning_rate": 1.6210200347257232e-07,
+      "loss": 0.494,
+      "step": 20241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5377411723066863,
+      "learning_rate": 1.6162818629118225e-07,
+      "loss": 0.7861,
+      "step": 20242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1868794154996745,
+      "learning_rate": 1.6115506035156991e-07,
+      "loss": 0.8047,
+      "step": 20243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9557300742020682,
+      "learning_rate": 1.6068262566689696e-07,
+      "loss": 0.8662,
+      "step": 20244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.409226298412856,
+      "learning_rate": 1.6021088225031676e-07,
+      "loss": 1.1318,
+      "step": 20245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7008524814778343,
+      "learning_rate": 1.5973983011495774e-07,
+      "loss": 0.998,
+      "step": 20246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1130241195305883,
+      "learning_rate": 1.5926946927392328e-07,
+      "loss": 0.8018,
+      "step": 20247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3441574759720734,
+      "learning_rate": 1.5879979974030844e-07,
+      "loss": 0.8232,
+      "step": 20248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2771914131851965,
+      "learning_rate": 1.5833082152718338e-07,
+      "loss": 0.6396,
+      "step": 20249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.361141716123296,
+      "learning_rate": 1.5786253464759594e-07,
+      "loss": 1.166,
+      "step": 20250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3457286366418164,
+      "learning_rate": 1.5739493911458292e-07,
+      "loss": 1.1338,
+      "step": 20251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3234403351061896,
+      "learning_rate": 1.5692803494115337e-07,
+      "loss": 1.1064,
+      "step": 20252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2675146181083288,
+      "learning_rate": 1.5646182214030246e-07,
+      "loss": 0.9653,
+      "step": 20253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7627614636995843,
+      "learning_rate": 1.559963007250087e-07,
+      "loss": 0.8242,
+      "step": 20254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8949963814693316,
+      "learning_rate": 1.5553147070822283e-07,
+      "loss": 1.1504,
+      "step": 20255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9048126304774868,
+      "learning_rate": 1.5506733210288172e-07,
+      "loss": 0.4683,
+      "step": 20256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.33118018411169,
+      "learning_rate": 1.5460388492190002e-07,
+      "loss": 1.9863,
+      "step": 20257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4465854222248535,
+      "learning_rate": 1.541411291781786e-07,
+      "loss": 1.0283,
+      "step": 20258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9334458315113188,
+      "learning_rate": 1.5367906488459592e-07,
+      "loss": 1.1094,
+      "step": 20259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.0204648186235685,
+      "learning_rate": 1.532176920540085e-07,
+      "loss": 1.4531,
+      "step": 20260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0822924366030877,
+      "learning_rate": 1.5275701069926152e-07,
+      "loss": 1.4219,
+      "step": 20261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.461013550567213,
+      "learning_rate": 1.52297020833167e-07,
+      "loss": 1.0703,
+      "step": 20262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.920384826844395,
+      "learning_rate": 1.5183772246853411e-07,
+      "loss": 1.0674,
+      "step": 20263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0514071673555416,
+      "learning_rate": 1.5137911561813877e-07,
+      "loss": 0.8945,
+      "step": 20264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.292541880084667,
+      "learning_rate": 1.5092120029474856e-07,
+      "loss": 1.293,
+      "step": 20265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3418462523789842,
+      "learning_rate": 1.50463976511106e-07,
+      "loss": 1.2988,
+      "step": 20266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.904317270886884,
+      "learning_rate": 1.500074442799343e-07,
+      "loss": 0.7446,
+      "step": 20267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.510305572266574,
+      "learning_rate": 1.4955160361393716e-07,
+      "loss": 1.2607,
+      "step": 20268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9008264284856278,
+      "learning_rate": 1.4909645452580444e-07,
+      "loss": 0.6069,
+      "step": 20269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.270628567017395,
+      "learning_rate": 1.4864199702819826e-07,
+      "loss": 0.3765,
+      "step": 20270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.622539711210685,
+      "learning_rate": 1.4818823113376955e-07,
+      "loss": 1.3027,
+      "step": 20271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.419853203317145,
+      "learning_rate": 1.4773515685514715e-07,
+      "loss": 1.1395,
+      "step": 20272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.235561429674571,
+      "learning_rate": 1.4728277420493487e-07,
+      "loss": 0.6842,
+      "step": 20273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.025760431900253,
+      "learning_rate": 1.4683108319572537e-07,
+      "loss": 1.2158,
+      "step": 20274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9855963485230799,
+      "learning_rate": 1.4638008384009194e-07,
+      "loss": 0.9646,
+      "step": 20275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.49867699902426,
+      "learning_rate": 1.459297761505829e-07,
+      "loss": 0.6919,
+      "step": 20276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3811872098559044,
+      "learning_rate": 1.4548016013972709e-07,
+      "loss": 1.1074,
+      "step": 20277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6054542707918023,
+      "learning_rate": 1.450312358200423e-07,
+      "loss": 0.6374,
+      "step": 20278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5942233469508262,
+      "learning_rate": 1.4458300320401853e-07,
+      "loss": 1.5059,
+      "step": 20279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.590857861358627,
+      "learning_rate": 1.4413546230413189e-07,
+      "loss": 1.2754,
+      "step": 20280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.03776460241356,
+      "learning_rate": 1.4368861313283632e-07,
+      "loss": 1.0703,
+      "step": 20281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5064333156600895,
+      "learning_rate": 1.4324245570256633e-07,
+      "loss": 0.9258,
+      "step": 20282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.473144097436413,
+      "learning_rate": 1.4279699002574254e-07,
+      "loss": 0.8198,
+      "step": 20283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.469235722527471,
+      "learning_rate": 1.423522161147578e-07,
+      "loss": 0.4053,
+      "step": 20284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5537050557114385,
+      "learning_rate": 1.4190813398199109e-07,
+      "loss": 0.5122,
+      "step": 20285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4860472419344,
+      "learning_rate": 1.4146474363979921e-07,
+      "loss": 0.9355,
+      "step": 20286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.522551200808313,
+      "learning_rate": 1.4102204510052786e-07,
+      "loss": 0.3109,
+      "step": 20287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.958953989543577,
+      "learning_rate": 1.4058003837648937e-07,
+      "loss": 0.5767,
+      "step": 20288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.230713028995335,
+      "learning_rate": 1.4013872347999058e-07,
+      "loss": 1.3369,
+      "step": 20289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9227688579296707,
+      "learning_rate": 1.396981004233078e-07,
+      "loss": 1.541,
+      "step": 20290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9333277047283628,
+      "learning_rate": 1.3925816921870893e-07,
+      "loss": 1.3965,
+      "step": 20291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7255993210270213,
+      "learning_rate": 1.3881892987843426e-07,
+      "loss": 1.2422,
+      "step": 20292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6176874791554865,
+      "learning_rate": 1.3838038241470729e-07,
+      "loss": 1.1504,
+      "step": 20293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.991057407491876,
+      "learning_rate": 1.379425268397322e-07,
+      "loss": 1.8867,
+      "step": 20294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.49067399775374,
+      "learning_rate": 1.3750536316569363e-07,
+      "loss": 0.7495,
+      "step": 20295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7339106574205325,
+      "learning_rate": 1.3706889140476242e-07,
+      "loss": 1.3027,
+      "step": 20296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.332233873826577,
+      "learning_rate": 1.3663311156908166e-07,
+      "loss": 0.8684,
+      "step": 20297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.177269918664035,
+      "learning_rate": 1.361980236707805e-07,
+      "loss": 1.1582,
+      "step": 20298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.269762817081348,
+      "learning_rate": 1.3576362772196315e-07,
+      "loss": 1.3643,
+      "step": 20299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4844018706781914,
+      "learning_rate": 1.3532992373472275e-07,
+      "loss": 1.0186,
+      "step": 20300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.481805356989961,
+      "learning_rate": 1.3489691172113017e-07,
+      "loss": 1.1094,
+      "step": 20301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.635724186786992,
+      "learning_rate": 1.344645916932341e-07,
+      "loss": 0.9697,
+      "step": 20302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.37230527881218,
+      "learning_rate": 1.3403296366306384e-07,
+      "loss": 1.4727,
+      "step": 20303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.673634277608791,
+      "learning_rate": 1.3360202764263474e-07,
+      "loss": 1.5703,
+      "step": 20304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.193563192320184,
+      "learning_rate": 1.3317178364393724e-07,
+      "loss": 1.1279,
+      "step": 20305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6128150056448973,
+      "learning_rate": 1.3274223167894506e-07,
+      "loss": 0.6201,
+      "step": 20306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0351063434829806,
+      "learning_rate": 1.3231337175961255e-07,
+      "loss": 0.9668,
+      "step": 20307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.357733036909838,
+      "learning_rate": 1.3188520389787463e-07,
+      "loss": 0.7764,
+      "step": 20308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.009468214730723,
+      "learning_rate": 1.3145772810565228e-07,
+      "loss": 1.2803,
+      "step": 20309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.945090285022376,
+      "learning_rate": 1.310309443948332e-07,
+      "loss": 0.7178,
+      "step": 20310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.872395286055498,
+      "learning_rate": 1.3060485277730238e-07,
+      "loss": 1.1182,
+      "step": 20311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.11497894849575,
+      "learning_rate": 1.3017945326490865e-07,
+      "loss": 1.0928,
+      "step": 20312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.350224561485483,
+      "learning_rate": 1.2975474586950088e-07,
+      "loss": 1.123,
+      "step": 20313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0281384624863157,
+      "learning_rate": 1.293307306028918e-07,
+      "loss": 1.3555,
+      "step": 20314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7070563509219108,
+      "learning_rate": 1.289074074768859e-07,
+      "loss": 1.3555,
+      "step": 20315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2397062282702693,
+      "learning_rate": 1.2848477650325984e-07,
+      "loss": 1.9492,
+      "step": 20316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.897548825853018,
+      "learning_rate": 1.2806283769377646e-07,
+      "loss": 1.9316,
+      "step": 20317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8009685203692687,
+      "learning_rate": 1.2764159106018193e-07,
+      "loss": 0.8359,
+      "step": 20318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.212475440975331,
+      "learning_rate": 1.272210366141946e-07,
+      "loss": 0.916,
+      "step": 20319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1722978139192604,
+      "learning_rate": 1.2680117436751904e-07,
+      "loss": 0.8662,
+      "step": 20320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.174762392348149,
+      "learning_rate": 1.2638200433184034e-07,
+      "loss": 1.4326,
+      "step": 20321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.887670708436356,
+      "learning_rate": 1.2596352651882693e-07,
+      "loss": 1.4707,
+      "step": 20322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.234046538425744,
+      "learning_rate": 1.2554574094012228e-07,
+      "loss": 0.916,
+      "step": 20323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.263207559962775,
+      "learning_rate": 1.2512864760735322e-07,
+      "loss": 1.3027,
+      "step": 20324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.431590234000791,
+      "learning_rate": 1.247122465321243e-07,
+      "loss": 1.0527,
+      "step": 20325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8034314974777526,
+      "learning_rate": 1.242965377260319e-07,
+      "loss": 0.8799,
+      "step": 20326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2117148606492867,
+      "learning_rate": 1.2388152120063612e-07,
+      "loss": 1.1797,
+      "step": 20327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7974347113582343,
+      "learning_rate": 1.2346719696749165e-07,
+      "loss": 1.7363,
+      "step": 20328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.123133364655197,
+      "learning_rate": 1.2305356503812537e-07,
+      "loss": 1.3281,
+      "step": 20329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4626735800353647,
+      "learning_rate": 1.2264062542405586e-07,
+      "loss": 0.7783,
+      "step": 20330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8887956759094817,
+      "learning_rate": 1.222283781367656e-07,
+      "loss": 1.2695,
+      "step": 20331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8596591495596386,
+      "learning_rate": 1.2181682318773424e-07,
+      "loss": 0.9912,
+      "step": 20332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0111659207026413,
+      "learning_rate": 1.214059605884138e-07,
+      "loss": 0.8694,
+      "step": 20333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7798865758961124,
+      "learning_rate": 1.20995790350234e-07,
+      "loss": 1.124,
+      "step": 20334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4113397644642034,
+      "learning_rate": 1.2058631248461628e-07,
+      "loss": 0.54,
+      "step": 20335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.744178818933328,
+      "learning_rate": 1.201775270029515e-07,
+      "loss": 0.7383,
+      "step": 20336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.413172485286259,
+      "learning_rate": 1.1976943391661665e-07,
+      "loss": 0.9807,
+      "step": 20337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.819205475499679,
+      "learning_rate": 1.1936203323697215e-07,
+      "loss": 0.3184,
+      "step": 20338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.722101585032411,
+      "learning_rate": 1.1895532497535334e-07,
+      "loss": 1.4443,
+      "step": 20339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.745445919405708,
+      "learning_rate": 1.1854930914307615e-07,
+      "loss": 1.1162,
+      "step": 20340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.212993407324526,
+      "learning_rate": 1.1814398575144548e-07,
+      "loss": 1.2241,
+      "step": 20341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.063089181284157,
+      "learning_rate": 1.177393548117356e-07,
+      "loss": 0.6226,
+      "step": 20342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7345330106222696,
+      "learning_rate": 1.173354163352125e-07,
+      "loss": 0.8623,
+      "step": 20343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.884664297422784,
+      "learning_rate": 1.1693217033311165e-07,
+      "loss": 0.7017,
+      "step": 20344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.3902378747051545,
+      "learning_rate": 1.165296168166602e-07,
+      "loss": 1.4766,
+      "step": 20345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.895875689817053,
+      "learning_rate": 1.1612775579706026e-07,
+      "loss": 0.9014,
+      "step": 20346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.13264126202639,
+      "learning_rate": 1.1572658728549457e-07,
+      "loss": 0.7085,
+      "step": 20347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4054412473498297,
+      "learning_rate": 1.1532611129312642e-07,
+      "loss": 0.6309,
+      "step": 20348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.689587159650244,
+      "learning_rate": 1.1492632783110247e-07,
+      "loss": 0.9531,
+      "step": 20349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.934664840869314,
+      "learning_rate": 1.145272369105499e-07,
+      "loss": 0.792,
+      "step": 20350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.781757602050789,
+      "learning_rate": 1.1412883854257095e-07,
+      "loss": 0.8789,
+      "step": 20351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.0979617440190195,
+      "learning_rate": 1.1373113273825675e-07,
+      "loss": 1.0352,
+      "step": 20352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.8162531894145575,
+      "learning_rate": 1.1333411950867346e-07,
+      "loss": 1.1445,
+      "step": 20353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.227704617597254,
+      "learning_rate": 1.1293779886487054e-07,
+      "loss": 1.0703,
+      "step": 20354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.367668581381479,
+      "learning_rate": 1.1254217081787533e-07,
+      "loss": 0.8726,
+      "step": 20355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.131064400217168,
+      "learning_rate": 1.1214723537870397e-07,
+      "loss": 1.4688,
+      "step": 20356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8703077094734557,
+      "learning_rate": 1.1175299255833938e-07,
+      "loss": 0.7479,
+      "step": 20357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.118741299321587,
+      "learning_rate": 1.1135944236775886e-07,
+      "loss": 1.1852,
+      "step": 20358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.993622517907714,
+      "learning_rate": 1.10966584817912e-07,
+      "loss": 1.5234,
+      "step": 20359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.246714099964381,
+      "learning_rate": 1.1057441991973173e-07,
+      "loss": 1.127,
+      "step": 20360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.947961562031605,
+      "learning_rate": 1.1018294768413706e-07,
+      "loss": 0.8506,
+      "step": 20361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.250772293830162,
+      "learning_rate": 1.0979216812201375e-07,
+      "loss": 1.1504,
+      "step": 20362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.806937927342552,
+      "learning_rate": 1.0940208124424477e-07,
+      "loss": 0.8867,
+      "step": 20363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.195460482070617,
+      "learning_rate": 1.0901268706167978e-07,
+      "loss": 1.1777,
+      "step": 20364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3088224617950113,
+      "learning_rate": 1.0862398558516006e-07,
+      "loss": 1.0781,
+      "step": 20365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.692553714368711,
+      "learning_rate": 1.0823597682550201e-07,
+      "loss": 0.7847,
+      "step": 20366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0485404424457765,
+      "learning_rate": 1.0784866079350253e-07,
+      "loss": 0.7811,
+      "step": 20367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.943542434275057,
+      "learning_rate": 1.0746203749994188e-07,
+      "loss": 0.8623,
+      "step": 20368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8740799235885215,
+      "learning_rate": 1.0707610695557813e-07,
+      "loss": 1.0303,
+      "step": 20369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.292613742856598,
+      "learning_rate": 1.0669086917115545e-07,
+      "loss": 0.9219,
+      "step": 20370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9218217458566897,
+      "learning_rate": 1.063063241573875e-07,
+      "loss": 1.1094,
+      "step": 20371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.143881404853046,
+      "learning_rate": 1.0592247192498239e-07,
+      "loss": 1.125,
+      "step": 20372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.696116982522692,
+      "learning_rate": 1.0553931248462046e-07,
+      "loss": 0.9072,
+      "step": 20373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.719314844543508,
+      "learning_rate": 1.051568458469654e-07,
+      "loss": 1.0918,
+      "step": 20374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0242724960300165,
+      "learning_rate": 1.0477507202265868e-07,
+      "loss": 0.5986,
+      "step": 20375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.795541584189773,
+      "learning_rate": 1.0439399102232794e-07,
+      "loss": 0.7334,
+      "step": 20376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.249325757199146,
+      "learning_rate": 1.0401360285657858e-07,
+      "loss": 0.5049,
+      "step": 20377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5476100305636202,
+      "learning_rate": 1.0363390753599377e-07,
+      "loss": 0.8203,
+      "step": 20378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7386300249864575,
+      "learning_rate": 1.0325490507114289e-07,
+      "loss": 1.1309,
+      "step": 20379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2760790088427623,
+      "learning_rate": 1.02876595472573e-07,
+      "loss": 0.8286,
+      "step": 20380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.217900169395817,
+      "learning_rate": 1.0249897875081183e-07,
+      "loss": 0.7637,
+      "step": 20381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.345294773815644,
+      "learning_rate": 1.0212205491636761e-07,
+      "loss": 0.5132,
+      "step": 20382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3714847650780544,
+      "learning_rate": 1.0174582397973198e-07,
+      "loss": 1.2559,
+      "step": 20383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.45404359791826,
+      "learning_rate": 1.0137028595137155e-07,
+      "loss": 1.5259,
+      "step": 20384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7219220653367073,
+      "learning_rate": 1.0099544084174184e-07,
+      "loss": 1.1204,
+      "step": 20385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1436686909106037,
+      "learning_rate": 1.0062128866127063e-07,
+      "loss": 1.0269,
+      "step": 20386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6625335870334665,
+      "learning_rate": 1.0024782942037458e-07,
+      "loss": 1.0215,
+      "step": 20387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.335592170229199,
+      "learning_rate": 9.987506312944261e-08,
+      "loss": 1.0879,
+      "step": 20388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9939761082390994,
+      "learning_rate": 9.950298979885254e-08,
+      "loss": 0.9495,
+      "step": 20389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7533059622517873,
+      "learning_rate": 9.913160943895716e-08,
+      "loss": 1.0957,
+      "step": 20390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.574009566172158,
+      "learning_rate": 9.876092206009269e-08,
+      "loss": 1.1436,
+      "step": 20391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.015183866266813,
+      "learning_rate": 9.839092767257308e-08,
+      "loss": 1.3711,
+      "step": 20392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.30366668567446,
+      "learning_rate": 9.802162628669565e-08,
+      "loss": 1.5059,
+      "step": 20393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2421625115999224,
+      "learning_rate": 9.765301791274107e-08,
+      "loss": 1.502,
+      "step": 20394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.742780485458571,
+      "learning_rate": 9.728510256096223e-08,
+      "loss": 1.4463,
+      "step": 20395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.2597872014387566,
+      "learning_rate": 9.691788024160376e-08,
+      "loss": 0.8813,
+      "step": 20396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9999660846695604,
+      "learning_rate": 9.655135096488244e-08,
+      "loss": 1.1016,
+      "step": 20397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.775881268207548,
+      "learning_rate": 9.618551474099846e-08,
+      "loss": 0.6274,
+      "step": 20398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8120018411904355,
+      "learning_rate": 9.582037158013258e-08,
+      "loss": 1.2754,
+      "step": 20399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4797592724377084,
+      "learning_rate": 9.545592149244609e-08,
+      "loss": 1.0713,
+      "step": 20400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1747808981846064,
+      "learning_rate": 9.509216448808367e-08,
+      "loss": 1.1416,
+      "step": 20401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0620242639440485,
+      "learning_rate": 9.472910057716777e-08,
+      "loss": 1.3105,
+      "step": 20402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.01430913648758,
+      "learning_rate": 9.436672976980421e-08,
+      "loss": 1.5547,
+      "step": 20403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.946125613697161,
+      "learning_rate": 9.400505207607379e-08,
+      "loss": 0.854,
+      "step": 20404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.556453375854241,
+      "learning_rate": 9.364406750604348e-08,
+      "loss": 1.0801,
+      "step": 20405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2129356601910097,
+      "learning_rate": 9.328377606975802e-08,
+      "loss": 1.209,
+      "step": 20406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.203681540247939,
+      "learning_rate": 9.292417777724827e-08,
+      "loss": 1.5527,
+      "step": 20407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.869854670094489,
+      "learning_rate": 9.256527263851733e-08,
+      "loss": 0.5952,
+      "step": 20408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9421286943244485,
+      "learning_rate": 9.220706066355445e-08,
+      "loss": 0.5586,
+      "step": 20409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.367476543241832,
+      "learning_rate": 9.184954186232942e-08,
+      "loss": 0.7852,
+      "step": 20410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.637009803814506,
+      "learning_rate": 9.149271624479261e-08,
+      "loss": 0.6025,
+      "step": 20411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8715366802896372,
+      "learning_rate": 9.11365838208722e-08,
+      "loss": 1.2656,
+      "step": 20412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3141374718822414,
+      "learning_rate": 9.078114460047971e-08,
+      "loss": 1.2373,
+      "step": 20413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.688534478381023,
+      "learning_rate": 9.042639859350998e-08,
+      "loss": 0.9199,
+      "step": 20414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.210711356415735,
+      "learning_rate": 9.007234580983014e-08,
+      "loss": 1.2188,
+      "step": 20415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.033986427038429,
+      "learning_rate": 8.971898625929898e-08,
+      "loss": 1.1484,
+      "step": 20416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.0068998431642155,
+      "learning_rate": 8.936631995174471e-08,
+      "loss": 1.8926,
+      "step": 20417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.447844148468037,
+      "learning_rate": 8.901434689698451e-08,
+      "loss": 1.165,
+      "step": 20418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9733989250179733,
+      "learning_rate": 8.866306710481609e-08,
+      "loss": 1.3789,
+      "step": 20419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8622398824653725,
+      "learning_rate": 8.831248058501218e-08,
+      "loss": 1.1406,
+      "step": 20420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.179912924703752,
+      "learning_rate": 8.796258734732887e-08,
+      "loss": 1.0576,
+      "step": 20421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0209396439212526,
+      "learning_rate": 8.76133874015056e-08,
+      "loss": 1.0186,
+      "step": 20422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5053526796457137,
+      "learning_rate": 8.726488075726235e-08,
+      "loss": 0.4021,
+      "step": 20423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.629029369421462,
+      "learning_rate": 8.691706742429417e-08,
+      "loss": 1.0889,
+      "step": 20424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.792090217761355,
+      "learning_rate": 8.656994741227942e-08,
+      "loss": 1.2051,
+      "step": 20425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2473773506144195,
+      "learning_rate": 8.622352073088257e-08,
+      "loss": 1.3379,
+      "step": 20426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0152053145663658,
+      "learning_rate": 8.587778738974317e-08,
+      "loss": 1.623,
+      "step": 20427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9306011430890897,
+      "learning_rate": 8.553274739848405e-08,
+      "loss": 1.0166,
+      "step": 20428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.504085777763598,
+      "learning_rate": 8.518840076670309e-08,
+      "loss": 0.4761,
+      "step": 20429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.163647715108979,
+      "learning_rate": 8.484474750398708e-08,
+      "loss": 1.2227,
+      "step": 20430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6293106126080372,
+      "learning_rate": 8.45017876198978e-08,
+      "loss": 0.8304,
+      "step": 20431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.651882125841322,
+      "learning_rate": 8.415952112398318e-08,
+      "loss": 1.0879,
+      "step": 20432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.096219705026113,
+      "learning_rate": 8.381794802576337e-08,
+      "loss": 0.9172,
+      "step": 20433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6924941724623366,
+      "learning_rate": 8.347706833474744e-08,
+      "loss": 0.9229,
+      "step": 20434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.075182442563971,
+      "learning_rate": 8.313688206042226e-08,
+      "loss": 0.8552,
+      "step": 20435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.34661469820261,
+      "learning_rate": 8.279738921225245e-08,
+      "loss": 1.293,
+      "step": 20436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.721590159038354,
+      "learning_rate": 8.245858979968878e-08,
+      "loss": 0.6187,
+      "step": 20437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.152113383264642,
+      "learning_rate": 8.212048383215709e-08,
+      "loss": 1.4258,
+      "step": 20438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7883099734856858,
+      "learning_rate": 8.178307131906926e-08,
+      "loss": 1.335,
+      "step": 20439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5655646276419675,
+      "learning_rate": 8.1446352269815e-08,
+      "loss": 0.5752,
+      "step": 20440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.316067495551763,
+      "learning_rate": 8.111032669376184e-08,
+      "loss": 1.0029,
+      "step": 20441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.922666310607633,
+      "learning_rate": 8.077499460026616e-08,
+      "loss": 1.208,
+      "step": 20442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.641611527412772,
+      "learning_rate": 8.044035599865662e-08,
+      "loss": 0.8936,
+      "step": 20443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.029317354327236,
+      "learning_rate": 8.0106410898248e-08,
+      "loss": 1.3281,
+      "step": 20444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.511106538969886,
+      "learning_rate": 7.977315930833285e-08,
+      "loss": 1.4883,
+      "step": 20445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.51235049869682,
+      "learning_rate": 7.944060123818708e-08,
+      "loss": 0.8855,
+      "step": 20446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4182745653671345,
+      "learning_rate": 7.910873669706165e-08,
+      "loss": 0.5187,
+      "step": 20447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2678852251200077,
+      "learning_rate": 7.877756569419636e-08,
+      "loss": 1.0527,
+      "step": 20448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.440195928018539,
+      "learning_rate": 7.844708823880609e-08,
+      "loss": 1.2012,
+      "step": 20449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5043452401459803,
+      "learning_rate": 7.811730434008624e-08,
+      "loss": 1.4668,
+      "step": 20450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.854043147004657,
+      "learning_rate": 7.778821400721837e-08,
+      "loss": 1.1738,
+      "step": 20451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.1206297561656635,
+      "learning_rate": 7.745981724935625e-08,
+      "loss": 1.4629,
+      "step": 20452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3695095755863735,
+      "learning_rate": 7.713211407564258e-08,
+      "loss": 0.9258,
+      "step": 20453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7624091293095727,
+      "learning_rate": 7.680510449519784e-08,
+      "loss": 0.6504,
+      "step": 20454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.202902816498996,
+      "learning_rate": 7.647878851711754e-08,
+      "loss": 0.9883,
+      "step": 20455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7634999464087486,
+      "learning_rate": 7.615316615048607e-08,
+      "loss": 0.9111,
+      "step": 20456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.296760014702838,
+      "learning_rate": 7.58282374043684e-08,
+      "loss": 0.8838,
+      "step": 20457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.955828146497591,
+      "learning_rate": 7.550400228780452e-08,
+      "loss": 0.4299,
+      "step": 20458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2983946671115656,
+      "learning_rate": 7.5180460809815e-08,
+      "loss": 0.7314,
+      "step": 20459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8144448231899593,
+      "learning_rate": 7.485761297940652e-08,
+      "loss": 1.0371,
+      "step": 20460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.584937233522551,
+      "learning_rate": 7.453545880556634e-08,
+      "loss": 0.8108,
+      "step": 20461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9745986289195105,
+      "learning_rate": 7.421399829725396e-08,
+      "loss": 0.7812,
+      "step": 20462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.411077635494113,
+      "learning_rate": 7.389323146342053e-08,
+      "loss": 1.1797,
+      "step": 20463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.14492821776822,
+      "learning_rate": 7.35731583129895e-08,
+      "loss": 1.4941,
+      "step": 20464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.425940898102663,
+      "learning_rate": 7.325377885487039e-08,
+      "loss": 1.0371,
+      "step": 20465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9468020611147496,
+      "learning_rate": 7.293509309795332e-08,
+      "loss": 0.9883,
+      "step": 20466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.454379867207604,
+      "learning_rate": 7.261710105110342e-08,
+      "loss": 1.1504,
+      "step": 20467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.974779471462672,
+      "learning_rate": 7.229980272316916e-08,
+      "loss": 1.2959,
+      "step": 20468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.173405981199426,
+      "learning_rate": 7.198319812298515e-08,
+      "loss": 0.5098,
+      "step": 20469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.490783487252775,
+      "learning_rate": 7.166728725936101e-08,
+      "loss": 1.3076,
+      "step": 20470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.980823071327102,
+      "learning_rate": 7.135207014108969e-08,
+      "loss": 1.0088,
+      "step": 20471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0258844963505966,
+      "learning_rate": 7.10375467769392e-08,
+      "loss": 0.7178,
+      "step": 20472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.6742221205512875,
+      "learning_rate": 7.07237171756664e-08,
+      "loss": 1.1484,
+      "step": 20473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.649212922494315,
+      "learning_rate": 7.04105813460032e-08,
+      "loss": 1.5654,
+      "step": 20474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.739906676690564,
+      "learning_rate": 7.009813929666209e-08,
+      "loss": 1.1758,
+      "step": 20475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3276733369949754,
+      "learning_rate": 6.978639103634444e-08,
+      "loss": 1.0801,
+      "step": 20476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.534979817640993,
+      "learning_rate": 6.947533657372107e-08,
+      "loss": 0.8467,
+      "step": 20477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2765690430087866,
+      "learning_rate": 6.916497591745175e-08,
+      "loss": 1.1992,
+      "step": 20478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.623690651798655,
+      "learning_rate": 6.885530907616844e-08,
+      "loss": 0.9297,
+      "step": 20479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.817789910802856,
+      "learning_rate": 6.854633605849481e-08,
+      "loss": 1.0381,
+      "step": 20480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.191555840865478,
+      "learning_rate": 6.823805687302676e-08,
+      "loss": 0.8857,
+      "step": 20481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.295275695796554,
+      "learning_rate": 6.793047152834353e-08,
+      "loss": 1.291,
+      "step": 20482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4391323028787513,
+      "learning_rate": 6.762358003300495e-08,
+      "loss": 1.7949,
+      "step": 20483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.402139099832449,
+      "learning_rate": 6.731738239555418e-08,
+      "loss": 1.584,
+      "step": 20484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.83447138295806,
+      "learning_rate": 6.701187862450941e-08,
+      "loss": 1.3418,
+      "step": 20485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.308330142566542,
+      "learning_rate": 6.670706872837218e-08,
+      "loss": 1.0254,
+      "step": 20486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.064002777715873,
+      "learning_rate": 6.640295271562735e-08,
+      "loss": 0.709,
+      "step": 20487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.305240163972624,
+      "learning_rate": 6.609953059474039e-08,
+      "loss": 1.0195,
+      "step": 20488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8685188175975935,
+      "learning_rate": 6.579680237415175e-08,
+      "loss": 0.6375,
+      "step": 20489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.072923017948901,
+      "learning_rate": 6.549476806228527e-08,
+      "loss": 1.3691,
+      "step": 20490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.888629781482101,
+      "learning_rate": 6.519342766755088e-08,
+      "loss": 0.8311,
+      "step": 20491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.673320802943036,
+      "learning_rate": 6.489278119833075e-08,
+      "loss": 1.248,
+      "step": 20492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.164053465896644,
+      "learning_rate": 6.459282866299598e-08,
+      "loss": 1.2676,
+      "step": 20493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7707765520822143,
+      "learning_rate": 6.429357006988712e-08,
+      "loss": 1.0918,
+      "step": 20494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.475233425345725,
+      "learning_rate": 6.399500542734194e-08,
+      "loss": 1.373,
+      "step": 20495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.252657698213882,
+      "learning_rate": 6.369713474366212e-08,
+      "loss": 1.4004,
+      "step": 20496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.11271042086184,
+      "learning_rate": 6.339995802713828e-08,
+      "loss": 0.9092,
+      "step": 20497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8311321646650955,
+      "learning_rate": 6.310347528604432e-08,
+      "loss": 0.6348,
+      "step": 20498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1219574421945637,
+      "learning_rate": 6.280768652862645e-08,
+      "loss": 1.5039,
+      "step": 20499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1825258649798753,
+      "learning_rate": 6.251259176311975e-08,
+      "loss": 0.6919,
+      "step": 20500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.748550997121714,
+      "learning_rate": 6.221819099773707e-08,
+      "loss": 0.9189,
+      "step": 20501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.431169139528008,
+      "learning_rate": 6.19244842406691e-08,
+      "loss": 1.1768,
+      "step": 20502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3725014374251305,
+      "learning_rate": 6.163147150008708e-08,
+      "loss": 1.1338,
+      "step": 20503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.458786572415631,
+      "learning_rate": 6.133915278415392e-08,
+      "loss": 0.8467,
+      "step": 20504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.213865998577548,
+      "learning_rate": 6.104752810099645e-08,
+      "loss": 1.291,
+      "step": 20505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.266203602538386,
+      "learning_rate": 6.075659745873596e-08,
+      "loss": 1.2207,
+      "step": 20506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6758070895764217,
+      "learning_rate": 6.04663608654632e-08,
+      "loss": 0.9312,
+      "step": 20507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9479804683035113,
+      "learning_rate": 6.017681832926059e-08,
+      "loss": 0.9771,
+      "step": 20508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6821768628091385,
+      "learning_rate": 5.988796985818557e-08,
+      "loss": 1.4355,
+      "step": 20509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4439525731594904,
+      "learning_rate": 5.959981546027338e-08,
+      "loss": 1.2617,
+      "step": 20510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.460539286048088,
+      "learning_rate": 5.931235514354538e-08,
+      "loss": 0.8643,
+      "step": 20511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3315324872964664,
+      "learning_rate": 5.902558891600074e-08,
+      "loss": 0.5732,
+      "step": 20512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0311431865946483,
+      "learning_rate": 5.873951678562195e-08,
+      "loss": 1.2197,
+      "step": 20513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.670367675896181,
+      "learning_rate": 5.845413876036654e-08,
+      "loss": 1.0645,
+      "step": 20514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7152264441938376,
+      "learning_rate": 5.8169454848180924e-08,
+      "loss": 0.6665,
+      "step": 20515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1658067204848885,
+      "learning_rate": 5.7885465056983775e-08,
+      "loss": 1.2637,
+      "step": 20516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.231029005887163,
+      "learning_rate": 5.760216939468266e-08,
+      "loss": 0.3992,
+      "step": 20517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6406345705840804,
+      "learning_rate": 5.731956786915737e-08,
+      "loss": 1.0742,
+      "step": 20518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.527584502832155,
+      "learning_rate": 5.703766048827386e-08,
+      "loss": 1.3262,
+      "step": 20519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.124518488984505,
+      "learning_rate": 5.675644725987861e-08,
+      "loss": 1.0684,
+      "step": 20520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.388372138671027,
+      "learning_rate": 5.64759281917987e-08,
+      "loss": 1.1406,
+      "step": 20521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.087405484738134,
+      "learning_rate": 5.619610329183622e-08,
+      "loss": 1.1055,
+      "step": 20522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.063239631813756,
+      "learning_rate": 5.591697256778494e-08,
+      "loss": 0.7705,
+      "step": 20523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0626208223137144,
+      "learning_rate": 5.563853602740532e-08,
+      "loss": 0.3987,
+      "step": 20524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.324141559411541,
+      "learning_rate": 5.536079367845226e-08,
+      "loss": 1.0,
+      "step": 20525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.060974558406944,
+      "learning_rate": 5.5083745528655696e-08,
+      "loss": 1.3149,
+      "step": 20526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3291352403553107,
+      "learning_rate": 5.480739158572057e-08,
+      "loss": 0.6272,
+      "step": 20527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.594712344724283,
+      "learning_rate": 5.453173185734073e-08,
+      "loss": 1.4688,
+      "step": 20528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0148637037958816,
+      "learning_rate": 5.425676635119059e-08,
+      "loss": 1.2461,
+      "step": 20529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.913231369496088,
+      "learning_rate": 5.398249507491682e-08,
+      "loss": 1.4004,
+      "step": 20530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.30491378737338,
+      "learning_rate": 5.370891803615219e-08,
+      "loss": 1.1973,
+      "step": 20531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.50571475128254,
+      "learning_rate": 5.3436035242515615e-08,
+      "loss": 0.6465,
+      "step": 20532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.97964075138507,
+      "learning_rate": 5.316384670159824e-08,
+      "loss": 1.8398,
+      "step": 20533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.734507666110721,
+      "learning_rate": 5.289235242097457e-08,
+      "loss": 1.2148,
+      "step": 20534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.424280342149049,
+      "learning_rate": 5.262155240820244e-08,
+      "loss": 1.5352,
+      "step": 20535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0744338778266833,
+      "learning_rate": 5.235144667081471e-08,
+      "loss": 1.4395,
+      "step": 20536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9738464073061333,
+      "learning_rate": 5.208203521633037e-08,
+      "loss": 0.8193,
+      "step": 20537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.3839853930682455,
+      "learning_rate": 5.181331805224898e-08,
+      "loss": 0.9922,
+      "step": 20538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.630508578274712,
+      "learning_rate": 5.1545295186042344e-08,
+      "loss": 1.1035,
+      "step": 20539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.648924518856623,
+      "learning_rate": 5.12779666251767e-08,
+      "loss": 1.4844,
+      "step": 20540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3888900544905467,
+      "learning_rate": 5.101133237708777e-08,
+      "loss": 0.751,
+      "step": 20541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0255347028672532,
+      "learning_rate": 5.074539244919463e-08,
+      "loss": 1.3057,
+      "step": 20542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.153818647616609,
+      "learning_rate": 5.0480146848902454e-08,
+      "loss": 1.0391,
+      "step": 20543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2313198827273784,
+      "learning_rate": 5.021559558359146e-08,
+      "loss": 1.1641,
+      "step": 20544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.281983793043142,
+      "learning_rate": 4.9951738660622415e-08,
+      "loss": 1.2773,
+      "step": 20545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9195970437524736,
+      "learning_rate": 4.968857608733945e-08,
+      "loss": 1.0117,
+      "step": 20546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1989072643465537,
+      "learning_rate": 4.9426107871067275e-08,
+      "loss": 1.4609,
+      "step": 20547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.342970858449264,
+      "learning_rate": 4.916433401910836e-08,
+      "loss": 1.1621,
+      "step": 20548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8375227939849714,
+      "learning_rate": 4.8903254538745776e-08,
+      "loss": 1.2705,
+      "step": 20549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4885363170498853,
+      "learning_rate": 4.864286943725149e-08,
+      "loss": 0.5911,
+      "step": 20550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.88377569046717,
+      "learning_rate": 4.838317872186693e-08,
+      "loss": 0.8759,
+      "step": 20551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.503254679203184,
+      "learning_rate": 4.8124182399822414e-08,
+      "loss": 1.1016,
+      "step": 20552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3201392523311006,
+      "learning_rate": 4.7865880478320526e-08,
+      "loss": 1.0742,
+      "step": 20553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.102644016636844,
+      "learning_rate": 4.76082729645555e-08,
+      "loss": 0.8809,
+      "step": 20554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0336582872640894,
+      "learning_rate": 4.735135986569661e-08,
+      "loss": 1.1814,
+      "step": 20555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2477118433455927,
+      "learning_rate": 4.7095141188888137e-08,
+      "loss": 1.332,
+      "step": 20556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.801640692856108,
+      "learning_rate": 4.683961694126326e-08,
+      "loss": 1.1367,
+      "step": 20557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.708001121295376,
+      "learning_rate": 4.658478712993297e-08,
+      "loss": 1.0967,
+      "step": 20558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.343708073718604,
+      "learning_rate": 4.6330651761991585e-08,
+      "loss": 1.1631,
+      "step": 20559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.858227953195985,
+      "learning_rate": 4.607721084450844e-08,
+      "loss": 1.1445,
+      "step": 20560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.6184881608428645,
+      "learning_rate": 4.5824464384539025e-08,
+      "loss": 1.1445,
+      "step": 20561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.717822053122262,
+      "learning_rate": 4.5572412389113805e-08,
+      "loss": 1.3154,
+      "step": 20562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.157490138697714,
+      "learning_rate": 4.53210548652494e-08,
+      "loss": 0.8701,
+      "step": 20563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.298945257661365,
+      "learning_rate": 4.507039181994299e-08,
+      "loss": 1.2598,
+      "step": 20564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5320197525643442,
+      "learning_rate": 4.482042326016678e-08,
+      "loss": 0.714,
+      "step": 20565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3868886634293744,
+      "learning_rate": 4.457114919287908e-08,
+      "loss": 0.6343,
+      "step": 20566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.617748245811408,
+      "learning_rate": 4.4322569625018793e-08,
+      "loss": 0.8975,
+      "step": 20567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4873949315367625,
+      "learning_rate": 4.407468456349983e-08,
+      "loss": 0.6006,
+      "step": 20568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.830622279672109,
+      "learning_rate": 4.382749401522501e-08,
+      "loss": 0.447,
+      "step": 20569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3962138551026673,
+      "learning_rate": 4.3580997987069385e-08,
+      "loss": 1.0713,
+      "step": 20570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.234329279691477,
+      "learning_rate": 4.3335196485894124e-08,
+      "loss": 1.1133,
+      "step": 20571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1672738814842063,
+      "learning_rate": 4.3090089518540987e-08,
+      "loss": 0.7891,
+      "step": 20572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.25730800602211,
+      "learning_rate": 4.284567709183229e-08,
+      "loss": 1.2207,
+      "step": 20573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.843459823944297,
+      "learning_rate": 4.2601959212568156e-08,
+      "loss": 0.9707,
+      "step": 20574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9222385713890182,
+      "learning_rate": 4.235893588752926e-08,
+      "loss": 0.4614,
+      "step": 20575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4171673705453256,
+      "learning_rate": 4.211660712348242e-08,
+      "loss": 1.2661,
+      "step": 20576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.912700917922756,
+      "learning_rate": 4.1874972927169464e-08,
+      "loss": 1.0264,
+      "step": 20577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1573819689462894,
+      "learning_rate": 4.163403330531834e-08,
+      "loss": 0.7021,
+      "step": 20578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.77637526869125,
+      "learning_rate": 4.139378826462925e-08,
+      "loss": 0.8604,
+      "step": 20579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3734171762908436,
+      "learning_rate": 4.115423781178851e-08,
+      "loss": 1.0898,
+      "step": 20580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8717533521655074,
+      "learning_rate": 4.091538195346856e-08,
+      "loss": 0.8384,
+      "step": 20581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3117430199887696,
+      "learning_rate": 4.067722069630853e-08,
+      "loss": 0.958,
+      "step": 20582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3954147526066576,
+      "learning_rate": 4.0439754046942e-08,
+      "loss": 1.4844,
+      "step": 20583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.0652398124778015,
+      "learning_rate": 4.020298201197759e-08,
+      "loss": 1.1416,
+      "step": 20584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.516000376114567,
+      "learning_rate": 3.996690459800167e-08,
+      "loss": 0.9629,
+      "step": 20585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7932548876389591,
+      "learning_rate": 3.973152181158679e-08,
+      "loss": 0.8193,
+      "step": 20586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1373901408745977,
+      "learning_rate": 3.9496833659280474e-08,
+      "loss": 0.9521,
+      "step": 20587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.4848156224077025,
+      "learning_rate": 3.926284014761639e-08,
+      "loss": 1.1934,
+      "step": 20588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0985537139289723,
+      "learning_rate": 3.902954128310876e-08,
+      "loss": 1.3926,
+      "step": 20589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7844750211927276,
+      "learning_rate": 3.879693707224408e-08,
+      "loss": 1.1777,
+      "step": 20590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8646590858571779,
+      "learning_rate": 3.8565027521500484e-08,
+      "loss": 0.6675,
+      "step": 20591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.083686785914729,
+      "learning_rate": 3.833381263732838e-08,
+      "loss": 1.166,
+      "step": 20592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1511609380751233,
+      "learning_rate": 3.810329242616706e-08,
+      "loss": 1.2246,
+      "step": 20593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.526190720323243,
+      "learning_rate": 3.7873466894428054e-08,
+      "loss": 1.4033,
+      "step": 20594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.712312271046997,
+      "learning_rate": 3.764433604850626e-08,
+      "loss": 0.7822,
+      "step": 20595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6373503656110557,
+      "learning_rate": 3.741589989478267e-08,
+      "loss": 0.5972,
+      "step": 20596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.289615915730178,
+      "learning_rate": 3.718815843961332e-08,
+      "loss": 1.0691,
+      "step": 20597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2793089438711056,
+      "learning_rate": 3.696111168933203e-08,
+      "loss": 1.0039,
+      "step": 20598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.232519749049624,
+      "learning_rate": 3.673475965026429e-08,
+      "loss": 1.3252,
+      "step": 20599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.065829089017005,
+      "learning_rate": 3.6509102328702304e-08,
+      "loss": 1.0312,
+      "step": 20600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.07250894597496,
+      "learning_rate": 3.628413973092992e-08,
+      "loss": 0.5283,
+      "step": 20601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0067374472106096,
+      "learning_rate": 3.6059871863208803e-08,
+      "loss": 1.292,
+      "step": 20602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2701282864925916,
+      "learning_rate": 3.5836298731778405e-08,
+      "loss": 1.2734,
+      "step": 20603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.890139317305915,
+      "learning_rate": 3.561342034285875e-08,
+      "loss": 0.9102,
+      "step": 20604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.592612419229213,
+      "learning_rate": 3.539123670265598e-08,
+      "loss": 1.7656,
+      "step": 20605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.182087291987875,
+      "learning_rate": 3.516974781735405e-08,
+      "loss": 0.8379,
+      "step": 20606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.667209142899933,
+      "learning_rate": 3.4948953693111906e-08,
+      "loss": 0.7676,
+      "step": 20607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.588815958727935,
+      "learning_rate": 3.472885433607742e-08,
+      "loss": 1.2852,
+      "step": 20608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.293732761115043,
+      "learning_rate": 3.4509449752376244e-08,
+      "loss": 1.4492,
+      "step": 20609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7725702714789913,
+      "learning_rate": 3.42907399481146e-08,
+      "loss": 0.6943,
+      "step": 20610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7326897346342645,
+      "learning_rate": 3.407272492937652e-08,
+      "loss": 0.5833,
+      "step": 20611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.008515116058432,
+      "learning_rate": 3.385540470223214e-08,
+      "loss": 1.0801,
+      "step": 20612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4311096347015533,
+      "learning_rate": 3.363877927272663e-08,
+      "loss": 1.1191,
+      "step": 20613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3843747880168134,
+      "learning_rate": 3.342284864689127e-08,
+      "loss": 1.4004,
+      "step": 20614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.200170462822082,
+      "learning_rate": 3.3207612830735144e-08,
+      "loss": 1.3242,
+      "step": 20615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.226544322408315,
+      "learning_rate": 3.2993071830245134e-08,
+      "loss": 1.1416,
+      "step": 20616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4292687093045413,
+      "learning_rate": 3.277922565139424e-08,
+      "loss": 1.166,
+      "step": 20617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9863900776356695,
+      "learning_rate": 3.256607430013325e-08,
+      "loss": 1.1118,
+      "step": 20618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.806083407459828,
+      "learning_rate": 3.235361778239632e-08,
+      "loss": 0.6152,
+      "step": 20619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.430326448356787,
+      "learning_rate": 3.214185610409259e-08,
+      "loss": 1.3438,
+      "step": 20620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1301329609475776,
+      "learning_rate": 3.193078927111459e-08,
+      "loss": 1.0234,
+      "step": 20621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9941369186501428,
+      "learning_rate": 3.172041728934094e-08,
+      "loss": 1.2109,
+      "step": 20622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7409640484397535,
+      "learning_rate": 3.151074016462252e-08,
+      "loss": 1.2217,
+      "step": 20623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.652873507420682,
+      "learning_rate": 3.1301757902793546e-08,
+      "loss": 0.4771,
+      "step": 20624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2068596967341025,
+      "learning_rate": 3.109347050967437e-08,
+      "loss": 0.9248,
+      "step": 20625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4130620272110637,
+      "learning_rate": 3.088587799105758e-08,
+      "loss": 0.8037,
+      "step": 20626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.687891554586291,
+      "learning_rate": 3.067898035272465e-08,
+      "loss": 0.5361,
+      "step": 20627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.437125837512268,
+      "learning_rate": 3.047277760042655e-08,
+      "loss": 0.8301,
+      "step": 20628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.064100860340695,
+      "learning_rate": 3.0267269739908676e-08,
+      "loss": 1.3223,
+      "step": 20629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.607147270915556,
+      "learning_rate": 3.006245677688591e-08,
+      "loss": 0.6499,
+      "step": 20630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.455556769974093,
+      "learning_rate": 2.9858338717062005e-08,
+      "loss": 0.8525,
+      "step": 20631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8285710947165903,
+      "learning_rate": 2.9654915566112994e-08,
+      "loss": 0.7539,
+      "step": 20632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.109765095924189,
+      "learning_rate": 2.9452187329703784e-08,
+      "loss": 1.4004,
+      "step": 20633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3515575320169178,
+      "learning_rate": 2.9250154013474306e-08,
+      "loss": 0.9219,
+      "step": 20634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.749781594952753,
+      "learning_rate": 2.904881562304784e-08,
+      "loss": 1.166,
+      "step": 20635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8620561213932185,
+      "learning_rate": 2.8848172164025465e-08,
+      "loss": 0.4355,
+      "step": 20636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5682766575405185,
+      "learning_rate": 2.8648223641994376e-08,
+      "loss": 1.5,
+      "step": 20637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7141515686627544,
+      "learning_rate": 2.8448970062516788e-08,
+      "loss": 1.1797,
+      "step": 20638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0449736405518184,
+      "learning_rate": 2.8250411431138268e-08,
+      "loss": 1.3213,
+      "step": 20639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.161171093587469,
+      "learning_rate": 2.805254775338495e-08,
+      "loss": 1.125,
+      "step": 20640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.0182320410799965,
+      "learning_rate": 2.785537903476354e-08,
+      "loss": 0.916,
+      "step": 20641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7684879553689112,
+      "learning_rate": 2.7658905280761315e-08,
+      "loss": 1.1523,
+      "step": 20642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.110051430927381,
+      "learning_rate": 2.746312649684335e-08,
+      "loss": 1.5156,
+      "step": 20643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.477220525676289,
+      "learning_rate": 2.726804268846084e-08,
+      "loss": 1.2578,
+      "step": 20644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.92079299651966,
+      "learning_rate": 2.707365386104277e-08,
+      "loss": 1.0645,
+      "step": 20645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.431477073284911,
+      "learning_rate": 2.6879960019998706e-08,
+      "loss": 0.6406,
+      "step": 20646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.071462789452639,
+      "learning_rate": 2.6686961170718783e-08,
+      "loss": 1.2266,
+      "step": 20647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6750444497666837,
+      "learning_rate": 2.64946573185737e-08,
+      "loss": 1.3027,
+      "step": 20648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.529685584875163,
+      "learning_rate": 2.6303048468917512e-08,
+      "loss": 1.0371,
+      "step": 20649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1024000349558647,
+      "learning_rate": 2.611213462707929e-08,
+      "loss": 1.332,
+      "step": 20650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.614232914962455,
+      "learning_rate": 2.5921915798371442e-08,
+      "loss": 0.7251,
+      "step": 20651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4027026892991326,
+      "learning_rate": 2.5732391988092518e-08,
+      "loss": 1.0132,
+      "step": 20652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.402097115610842,
+      "learning_rate": 2.5543563201513297e-08,
+      "loss": 1.1543,
+      "step": 20653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.432327792180735,
+      "learning_rate": 2.5355429443887913e-08,
+      "loss": 1.9844,
+      "step": 20654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.523260813098878,
+      "learning_rate": 2.5167990720456613e-08,
+      "loss": 0.8262,
+      "step": 20655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.494511874655611,
+      "learning_rate": 2.49812470364319e-08,
+      "loss": 0.5815,
+      "step": 20656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.331704247752611,
+      "learning_rate": 2.479519839700961e-08,
+      "loss": 2.1484,
+      "step": 20657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.370173380399897,
+      "learning_rate": 2.4609844807371718e-08,
+      "loss": 0.8667,
+      "step": 20658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.113106486968691,
+      "learning_rate": 2.4425186272675204e-08,
+      "loss": 1.6602,
+      "step": 20659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.700876790924597,
+      "learning_rate": 2.4241222798054852e-08,
+      "loss": 1.123,
+      "step": 20660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.97633372952067,
+      "learning_rate": 2.4057954388637117e-08,
+      "loss": 1.1758,
+      "step": 20661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.267554572982771,
+      "learning_rate": 2.387538104951792e-08,
+      "loss": 1.918,
+      "step": 20662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4185505353511054,
+      "learning_rate": 2.3693502785776534e-08,
+      "loss": 0.7944,
+      "step": 20663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.259870798323936,
+      "learning_rate": 2.3512319602481127e-08,
+      "loss": 1.6035,
+      "step": 20664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.674741414770519,
+      "learning_rate": 2.3331831504666557e-08,
+      "loss": 1.3311,
+      "step": 20665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7904440337380514,
+      "learning_rate": 2.315203849736214e-08,
+      "loss": 0.7168,
+      "step": 20666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4253322782777236,
+      "learning_rate": 2.297294058556665e-08,
+      "loss": 0.4995,
+      "step": 20667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.72489332600738,
+      "learning_rate": 2.279453777426499e-08,
+      "loss": 1.1123,
+      "step": 20668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1237990551256973,
+      "learning_rate": 2.261683006842541e-08,
+      "loss": 0.8477,
+      "step": 20669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.509970259115784,
+      "learning_rate": 2.24398174729884e-08,
+      "loss": 1.2832,
+      "step": 20670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.634982258617292,
+      "learning_rate": 2.2263499992886127e-08,
+      "loss": 1.1338,
+      "step": 20671
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.891776190743937,
+      "learning_rate": 2.2087877633020226e-08,
+      "loss": 0.7847,
+      "step": 20672
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.797204817079091,
+      "learning_rate": 2.1912950398278455e-08,
+      "loss": 0.8145,
+      "step": 20673
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.7892779602072535,
+      "learning_rate": 2.173871829353469e-08,
+      "loss": 1.3574,
+      "step": 20674
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1145056599099035,
+      "learning_rate": 2.1565181323632277e-08,
+      "loss": 0.9629,
+      "step": 20675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.325598092128881,
+      "learning_rate": 2.139233949340069e-08,
+      "loss": 1.0195,
+      "step": 20676
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.570786058668627,
+      "learning_rate": 2.122019280765275e-08,
+      "loss": 0.9707,
+      "step": 20677
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3904162546412038,
+      "learning_rate": 2.1048741271179063e-08,
+      "loss": 1.2363,
+      "step": 20678
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1316453086472555,
+      "learning_rate": 2.0877984888748038e-08,
+      "loss": 1.0889,
+      "step": 20679
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1159296772554215,
+      "learning_rate": 2.0707923665114203e-08,
+      "loss": 0.6396,
+      "step": 20680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.6619389746168665,
+      "learning_rate": 2.053855760500989e-08,
+      "loss": 0.9648,
+      "step": 20681
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.062083061537285,
+      "learning_rate": 2.036988671314799e-08,
+      "loss": 1.0498,
+      "step": 20682
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.431708179439382,
+      "learning_rate": 2.0201910994221974e-08,
+      "loss": 1.0723,
+      "step": 20683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.166014560336525,
+      "learning_rate": 2.003463045291143e-08,
+      "loss": 0.8535,
+      "step": 20684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6789134161478683,
+      "learning_rate": 1.9868045093862643e-08,
+      "loss": 1.4746,
+      "step": 20685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.448604334609793,
+      "learning_rate": 1.9702154921719117e-08,
+      "loss": 1.002,
+      "step": 20686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.165346216143599,
+      "learning_rate": 1.953695994109661e-08,
+      "loss": 0.9717,
+      "step": 20687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7389481765734667,
+      "learning_rate": 1.937246015658867e-08,
+      "loss": 0.519,
+      "step": 20688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.061877879325473,
+      "learning_rate": 1.9208655572777734e-08,
+      "loss": 1.3604,
+      "step": 20689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0515340542969027,
+      "learning_rate": 1.90455461942185e-08,
+      "loss": 0.4688,
+      "step": 20690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.617024521948586,
+      "learning_rate": 1.888313202545178e-08,
+      "loss": 1.0362,
+      "step": 20691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1952073693132292,
+      "learning_rate": 1.872141307099895e-08,
+      "loss": 0.9004,
+      "step": 20692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2676766261584222,
+      "learning_rate": 1.8560389335359197e-08,
+      "loss": 0.7578,
+      "step": 20693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.102555251995849,
+      "learning_rate": 1.8400060823012266e-08,
+      "loss": 1.75,
+      "step": 20694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0575270377291104,
+      "learning_rate": 1.824042753842403e-08,
+      "loss": 0.5654,
+      "step": 20695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8842785457894908,
+      "learning_rate": 1.808148948603261e-08,
+      "loss": 0.7744,
+      "step": 20696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.035371560332619,
+      "learning_rate": 1.7923246670265016e-08,
+      "loss": 0.8091,
+      "step": 20697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4789636566218656,
+      "learning_rate": 1.7765699095520506e-08,
+      "loss": 1.4043,
+      "step": 20698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.383904289619223,
+      "learning_rate": 1.760884676619001e-08,
+      "loss": 1.0886,
+      "step": 20699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.377770063302587,
+      "learning_rate": 1.7452689686631163e-08,
+      "loss": 1.084,
+      "step": 20700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1740865279423676,
+      "learning_rate": 1.7297227861196032e-08,
+      "loss": 0.6989,
+      "step": 20701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.064772405093,
+      "learning_rate": 1.7142461294208934e-08,
+      "loss": 1.373,
+      "step": 20702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4768871010410853,
+      "learning_rate": 1.698838998997476e-08,
+      "loss": 1.7227,
+      "step": 20703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4254071473934933,
+      "learning_rate": 1.683501395278453e-08,
+      "loss": 0.5693,
+      "step": 20704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.351592260124825,
+      "learning_rate": 1.6682333186904266e-08,
+      "loss": 0.8359,
+      "step": 20705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.70311565066799,
+      "learning_rate": 1.6530347696583347e-08,
+      "loss": 0.9106,
+      "step": 20706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.581462470748857,
+      "learning_rate": 1.6379057486054507e-08,
+      "loss": 1.0469,
+      "step": 20707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.086835790666537,
+      "learning_rate": 1.6228462559522704e-08,
+      "loss": 1.1699,
+      "step": 20708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.841441296751228,
+      "learning_rate": 1.607856292118459e-08,
+      "loss": 1.2793,
+      "step": 20709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6130299810885966,
+      "learning_rate": 1.5929358575206275e-08,
+      "loss": 0.4438,
+      "step": 20710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1624424788922876,
+      "learning_rate": 1.5780849525745544e-08,
+      "loss": 0.5503,
+      "step": 20711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.32449652029767,
+      "learning_rate": 1.5633035776932424e-08,
+      "loss": 1.0781,
+      "step": 20712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.246447314791209,
+      "learning_rate": 1.5485917332880296e-08,
+      "loss": 1.2793,
+      "step": 20713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4467500111505935,
+      "learning_rate": 1.53394941976831e-08,
+      "loss": 0.5176,
+      "step": 20714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.289444023582562,
+      "learning_rate": 1.519376637541814e-08,
+      "loss": 1.123,
+      "step": 20715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.952273672960176,
+      "learning_rate": 1.5048733870137722e-08,
+      "loss": 1.0254,
+      "step": 20716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.116371449515051,
+      "learning_rate": 1.4904396685880283e-08,
+      "loss": 0.6064,
+      "step": 20717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.460051279842334,
+      "learning_rate": 1.476075482666206e-08,
+      "loss": 0.3677,
+      "step": 20718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8397678538259155,
+      "learning_rate": 1.4617808296479851e-08,
+      "loss": 1.3252,
+      "step": 20719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.063969984154203,
+      "learning_rate": 1.4475557099313809e-08,
+      "loss": 1.1523,
+      "step": 20720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.009377329294921,
+      "learning_rate": 1.43340012391191e-08,
+      "loss": 1.3079,
+      "step": 20721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.217889893058533,
+      "learning_rate": 1.4193140719839793e-08,
+      "loss": 1.8125,
+      "step": 20722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2149437506704395,
+      "learning_rate": 1.40529755453922e-08,
+      "loss": 0.4502,
+      "step": 20723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5395297984947187,
+      "learning_rate": 1.3913505719678755e-08,
+      "loss": 0.856,
+      "step": 20724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.034313423639829,
+      "learning_rate": 1.3774731246582462e-08,
+      "loss": 0.755,
+      "step": 20725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9767690558180218,
+      "learning_rate": 1.3636652129961347e-08,
+      "loss": 0.686,
+      "step": 20726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2298796000052055,
+      "learning_rate": 1.3499268373662333e-08,
+      "loss": 0.7979,
+      "step": 20727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2222146934805598,
+      "learning_rate": 1.3362579981504586e-08,
+      "loss": 0.814,
+      "step": 20728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5286212033382784,
+      "learning_rate": 1.3226586957296173e-08,
+      "loss": 0.9736,
+      "step": 20729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.343409704299658,
+      "learning_rate": 1.30912893048174e-08,
+      "loss": 1.1821,
+      "step": 20730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.465323752478698,
+      "learning_rate": 1.2956687027837477e-08,
+      "loss": 1.0781,
+      "step": 20731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7072545079235915,
+      "learning_rate": 1.282278013010063e-08,
+      "loss": 1.2266,
+      "step": 20732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.577113779094317,
+      "learning_rate": 1.268956861533166e-08,
+      "loss": 0.6265,
+      "step": 20733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6523469860298743,
+      "learning_rate": 1.2557052487241483e-08,
+      "loss": 0.7236,
+      "step": 20734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7778345240899913,
+      "learning_rate": 1.2425231749516042e-08,
+      "loss": 0.5815,
+      "step": 20735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.058656439954043,
+      "learning_rate": 1.2294106405821848e-08,
+      "loss": 1.3242,
+      "step": 20736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.235708758489537,
+      "learning_rate": 1.216367645981431e-08,
+      "loss": 1.1299,
+      "step": 20737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9593383834313443,
+      "learning_rate": 1.2033941915115531e-08,
+      "loss": 0.9123,
+      "step": 20738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.978531153904191,
+      "learning_rate": 1.1904902775342064e-08,
+      "loss": 0.7695,
+      "step": 20739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3694281218864557,
+      "learning_rate": 1.1776559044079927e-08,
+      "loss": 1.2051,
+      "step": 20740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8439704397615047,
+      "learning_rate": 1.1648910724904038e-08,
+      "loss": 0.4644,
+      "step": 20741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6854087445890626,
+      "learning_rate": 1.1521957821367113e-08,
+      "loss": 0.5083,
+      "step": 20742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7774323396999843,
+      "learning_rate": 1.1395700337002434e-08,
+      "loss": 1.0361,
+      "step": 20743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3594070787894266,
+      "learning_rate": 1.1270138275321085e-08,
+      "loss": 0.3624,
+      "step": 20744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9681660830733525,
+      "learning_rate": 1.1145271639817489e-08,
+      "loss": 0.5095,
+      "step": 20745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.647637710663967,
+      "learning_rate": 1.1021100433969423e-08,
+      "loss": 1.2578,
+      "step": 20746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.672029402712784,
+      "learning_rate": 1.0897624661232452e-08,
+      "loss": 0.9453,
+      "step": 20747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3264259436647023,
+      "learning_rate": 1.0774844325039946e-08,
+      "loss": 0.5803,
+      "step": 20748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.973827479537723,
+      "learning_rate": 1.0652759428811388e-08,
+      "loss": 1.3809,
+      "step": 20749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.861136305697062,
+      "learning_rate": 1.0531369975941285e-08,
+      "loss": 0.48,
+      "step": 20750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.572329449442404,
+      "learning_rate": 1.0410675969813043e-08,
+      "loss": 1.2441,
+      "step": 20751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7703651573149553,
+      "learning_rate": 1.0290677413779538e-08,
+      "loss": 0.9102,
+      "step": 20752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.000765012064904,
+      "learning_rate": 1.017137431118531e-08,
+      "loss": 0.4517,
+      "step": 20753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7901315216457188,
+      "learning_rate": 1.0052766665347157e-08,
+      "loss": 1.5273,
+      "step": 20754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.987498780573512,
+      "learning_rate": 9.93485447956799e-09,
+      "loss": 1.5254,
+      "step": 20755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5389788335073398,
+      "learning_rate": 9.817637757125741e-09,
+      "loss": 1.1855,
+      "step": 20756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.767768352859472,
+      "learning_rate": 9.701116501287245e-09,
+      "loss": 0.9248,
+      "step": 20757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.426662467028505,
+      "learning_rate": 9.585290715294349e-09,
+      "loss": 1.7656,
+      "step": 20758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6313264994249437,
+      "learning_rate": 9.4701604023667e-09,
+      "loss": 1.2656,
+      "step": 20759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.87415593540121,
+      "learning_rate": 9.355725565712847e-09,
+      "loss": 0.4392,
+      "step": 20760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.363384654407471,
+      "learning_rate": 9.241986208513576e-09,
+      "loss": 1.2773,
+      "step": 20761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.153149734566405,
+      "learning_rate": 9.128942333938572e-09,
+      "loss": 1.1973,
+      "step": 20762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9561300734544906,
+      "learning_rate": 9.01659394512977e-09,
+      "loss": 1.1348,
+      "step": 20763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8136300477547835,
+      "learning_rate": 8.90494104521522e-09,
+      "loss": 1.002,
+      "step": 20764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6624232938369845,
+      "learning_rate": 8.793983637303548e-09,
+      "loss": 1.1133,
+      "step": 20765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1500015391854386,
+      "learning_rate": 8.6837217244784e-09,
+      "loss": 1.3242,
+      "step": 20766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.174767304173259,
+      "learning_rate": 8.574155309815091e-09,
+      "loss": 1.0093,
+      "step": 20767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.23121658396459,
+      "learning_rate": 8.465284396355632e-09,
+      "loss": 1.3105,
+      "step": 20768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4181512569639634,
+      "learning_rate": 8.357108987136486e-09,
+      "loss": 0.9316,
+      "step": 20769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.536867948091399,
+      "learning_rate": 8.24962908516358e-09,
+      "loss": 1.0303,
+      "step": 20770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3484334309860584,
+      "learning_rate": 8.142844693428964e-09,
+      "loss": 1.5796,
+      "step": 20771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.740310465959613,
+      "learning_rate": 8.036755814902485e-09,
+      "loss": 1.7266,
+      "step": 20772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.069258864781153,
+      "learning_rate": 7.93136245254289e-09,
+      "loss": 1.7793,
+      "step": 20773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5119677193323025,
+      "learning_rate": 7.826664609275613e-09,
+      "loss": 1.1328,
+      "step": 20774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8008698216168764,
+      "learning_rate": 7.722662288020543e-09,
+      "loss": 1.1064,
+      "step": 20775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.437402087470313,
+      "learning_rate": 7.619355491669811e-09,
+      "loss": 0.7739,
+      "step": 20776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7172966064869937,
+      "learning_rate": 7.516744223096117e-09,
+      "loss": 1.3574,
+      "step": 20777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0906491447844937,
+      "learning_rate": 7.4148284851582875e-09,
+      "loss": 1.1211,
+      "step": 20778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4068054945188333,
+      "learning_rate": 7.313608280690165e-09,
+      "loss": 0.792,
+      "step": 20779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1217279487184033,
+      "learning_rate": 7.213083612511717e-09,
+      "loss": 0.7776,
+      "step": 20780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.968099577501212,
+      "learning_rate": 7.113254483415155e-09,
+      "loss": 1.0449,
+      "step": 20781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8113558874250593,
+      "learning_rate": 7.014120896184362e-09,
+      "loss": 1.4785,
+      "step": 20782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5677972861382674,
+      "learning_rate": 6.915682853575467e-09,
+      "loss": 0.6006,
+      "step": 20783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.706457740615859,
+      "learning_rate": 6.8179403583307215e-09,
+      "loss": 0.6914,
+      "step": 20784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6704762684807366,
+      "learning_rate": 6.720893413164619e-09,
+      "loss": 1.332,
+      "step": 20785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.52462013333233,
+      "learning_rate": 6.624542020783331e-09,
+      "loss": 1.6406,
+      "step": 20786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.095041318878274,
+      "learning_rate": 6.528886183865268e-09,
+      "loss": 0.8262,
+      "step": 20787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.050061477382912,
+      "learning_rate": 6.433925905074967e-09,
+      "loss": 1.0645,
+      "step": 20788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.130309185065133,
+      "learning_rate": 6.339661187051982e-09,
+      "loss": 1.2559,
+      "step": 20789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2869927224267377,
+      "learning_rate": 6.246092032421991e-09,
+      "loss": 0.9629,
+      "step": 20790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4827122912341095,
+      "learning_rate": 6.1532184437884666e-09,
+      "loss": 0.5344,
+      "step": 20791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6451856966941123,
+      "learning_rate": 6.061040423735453e-09,
+      "loss": 0.9922,
+      "step": 20792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.668431433217226,
+      "learning_rate": 5.969557974827567e-09,
+      "loss": 1.4883,
+      "step": 20793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.183249837635839,
+      "learning_rate": 5.878771099612768e-09,
+      "loss": 1.3613,
+      "step": 20794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3257395152741664,
+      "learning_rate": 5.788679800616814e-09,
+      "loss": 1.2427,
+      "step": 20795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.044276990556257,
+      "learning_rate": 5.699284080346035e-09,
+      "loss": 1.0918,
+      "step": 20796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1754876697202254,
+      "learning_rate": 5.610583941290104e-09,
+      "loss": 0.6606,
+      "step": 20797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.746725051081069,
+      "learning_rate": 5.522579385916493e-09,
+      "loss": 1.1094,
+      "step": 20798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.252740991806597,
+      "learning_rate": 5.4352704166732436e-09,
+      "loss": 1.418,
+      "step": 20799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9077142813578964,
+      "learning_rate": 5.348657035991745e-09,
+      "loss": 0.7236,
+      "step": 20800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8698371378718834,
+      "learning_rate": 5.262739246281179e-09,
+      "loss": 0.5747,
+      "step": 20801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9571758925262426,
+      "learning_rate": 5.177517049934078e-09,
+      "loss": 1.3242,
+      "step": 20802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.633477766315222,
+      "learning_rate": 5.092990449320767e-09,
+      "loss": 1.3389,
+      "step": 20803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9813890594875008,
+      "learning_rate": 5.00915944679492e-09,
+      "loss": 1.6543,
+      "step": 20804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.895659463248517,
+      "learning_rate": 4.926024044688005e-09,
+      "loss": 1.1738,
+      "step": 20805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3924000451903535,
+      "learning_rate": 4.843584245314836e-09,
+      "loss": 1.2891,
+      "step": 20806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5266452867749134,
+      "learning_rate": 4.761840050968025e-09,
+      "loss": 0.563,
+      "step": 20807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.034644634107243,
+      "learning_rate": 4.680791463923529e-09,
+      "loss": 0.7581,
+      "step": 20808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.696634515712915,
+      "learning_rate": 4.600438486437875e-09,
+      "loss": 1.4531,
+      "step": 20809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3234850839489485,
+      "learning_rate": 4.520781120748163e-09,
+      "loss": 0.9131,
+      "step": 20810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0791181901416476,
+      "learning_rate": 4.441819369066514e-09,
+      "loss": 1.1825,
+      "step": 20811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0579150385907954,
+      "learning_rate": 4.363553233593942e-09,
+      "loss": 1.3516,
+      "step": 20812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8063195880511436,
+      "learning_rate": 4.2859827165092625e-09,
+      "loss": 0.9893,
+      "step": 20813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8877491434786404,
+      "learning_rate": 4.2091078199690824e-09,
+      "loss": 0.9502,
+      "step": 20814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.449758807803555,
+      "learning_rate": 4.132928546113357e-09,
+      "loss": 0.8223,
+      "step": 20815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5622029766272294,
+      "learning_rate": 4.057444897062612e-09,
+      "loss": 1.5508,
+      "step": 20816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1609060265514275,
+      "learning_rate": 3.982656874917945e-09,
+      "loss": 1.1738,
+      "step": 20817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.868843458446754,
+      "learning_rate": 3.908564481758248e-09,
+      "loss": 1.5479,
+      "step": 20818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.795696697835149,
+      "learning_rate": 3.835167719648536e-09,
+      "loss": 1.9805,
+      "step": 20819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.133446791234563,
+      "learning_rate": 3.7624665906316194e-09,
+      "loss": 1.4336,
+      "step": 20820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.397578778449386,
+      "learning_rate": 3.690461096728104e-09,
+      "loss": 1.0098,
+      "step": 20821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.198826516928593,
+      "learning_rate": 3.6191512399419427e-09,
+      "loss": 1.668,
+      "step": 20822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0679237449847907,
+      "learning_rate": 3.548537022260434e-09,
+      "loss": 0.4395,
+      "step": 20823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.699868123253733,
+      "learning_rate": 3.478618445645898e-09,
+      "loss": 1.1807,
+      "step": 20824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.359938295333584,
+      "learning_rate": 3.4093955120467756e-09,
+      "loss": 0.6672,
+      "step": 20825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.685578124818184,
+      "learning_rate": 3.3408682233865285e-09,
+      "loss": 1.3027,
+      "step": 20826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.4793819293342425,
+      "learning_rate": 3.27303658157474e-09,
+      "loss": 1.7383,
+      "step": 20827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.004685398841,
+      "learning_rate": 3.205900588496014e-09,
+      "loss": 0.8906,
+      "step": 20828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2797331274014576,
+      "learning_rate": 3.139460246023851e-09,
+      "loss": 1.3145,
+      "step": 20829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0020079250955396,
+      "learning_rate": 3.073715556001222e-09,
+      "loss": 1.0527,
+      "step": 20830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.019694956123313,
+      "learning_rate": 3.008666520262771e-09,
+      "loss": 0.8843,
+      "step": 20831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.397476397803306,
+      "learning_rate": 2.94431314061816e-09,
+      "loss": 1.0645,
+      "step": 20832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.28645504290113,
+      "learning_rate": 2.8806554188548496e-09,
+      "loss": 0.5791,
+      "step": 20833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5220183642548424,
+      "learning_rate": 2.8176933567464204e-09,
+      "loss": 1.1309,
+      "step": 20834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9522505176901332,
+      "learning_rate": 2.7554269560470246e-09,
+      "loss": 0.7695,
+      "step": 20835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6568845540601407,
+      "learning_rate": 2.6938562184858353e-09,
+      "loss": 1.2676,
+      "step": 20836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.031316948859651,
+      "learning_rate": 2.6329811457809215e-09,
+      "loss": 1.0166,
+      "step": 20837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.379161501565416,
+      "learning_rate": 2.5728017396198236e-09,
+      "loss": 0.9092,
+      "step": 20838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8503790037202883,
+      "learning_rate": 2.513318001684528e-09,
+      "loss": 0.6245,
+      "step": 20839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.266846646049436,
+      "learning_rate": 2.4545299336264927e-09,
+      "loss": 0.5417,
+      "step": 20840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.193439485576474,
+      "learning_rate": 2.3964375370805205e-09,
+      "loss": 1.2275,
+      "step": 20841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9523869959744475,
+      "learning_rate": 2.3390408136675368e-09,
+      "loss": 1.3027,
+      "step": 20842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2854357426234695,
+      "learning_rate": 2.282339764980712e-09,
+      "loss": 0.8848,
+      "step": 20843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.194147635827826,
+      "learning_rate": 2.2263343925993386e-09,
+      "loss": 1.4023,
+      "step": 20844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.406845810019467,
+      "learning_rate": 2.171024698086055e-09,
+      "loss": 1.377,
+      "step": 20845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2198269718192294,
+      "learning_rate": 2.116410682972969e-09,
+      "loss": 1.0068,
+      "step": 20846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9653907454350703,
+      "learning_rate": 2.062492348783862e-09,
+      "loss": 0.9358,
+      "step": 20847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0309713319213607,
+      "learning_rate": 2.0092696970203105e-09,
+      "loss": 1.0352,
+      "step": 20848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6169882627347145,
+      "learning_rate": 1.9567427291616867e-09,
+      "loss": 1.1641,
+      "step": 20849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8116123070141823,
+      "learning_rate": 1.9049114466707095e-09,
+      "loss": 1.541,
+      "step": 20850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.827674619104164,
+      "learning_rate": 1.8537758509878932e-09,
+      "loss": 0.8096,
+      "step": 20851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1313384275466007,
+      "learning_rate": 1.803335943537099e-09,
+      "loss": 0.4907,
+      "step": 20852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.484072108968198,
+      "learning_rate": 1.753591725725534e-09,
+      "loss": 1.1055,
+      "step": 20853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.852022526565235,
+      "learning_rate": 1.704543198932651e-09,
+      "loss": 0.6855,
+      "step": 20854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.505407379680994,
+      "learning_rate": 1.6561903645267995e-09,
+      "loss": 1.061,
+      "step": 20855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2097804408380046,
+      "learning_rate": 1.6085332238513495e-09,
+      "loss": 1.0996,
+      "step": 20856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1703280421580033,
+      "learning_rate": 1.5615717782330174e-09,
+      "loss": 0.5276,
+      "step": 20857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6096643527291072,
+      "learning_rate": 1.5153060289818665e-09,
+      "loss": 1.2578,
+      "step": 20858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6039140400739966,
+      "learning_rate": 1.4697359773802045e-09,
+      "loss": 0.6367,
+      "step": 20859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6049764874471686,
+      "learning_rate": 1.4248616246992364e-09,
+      "loss": 1.207,
+      "step": 20860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.0014685044111795,
+      "learning_rate": 1.3806829721879633e-09,
+      "loss": 1.3745,
+      "step": 20861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7211966279917603,
+      "learning_rate": 1.3372000210759572e-09,
+      "loss": 0.6724,
+      "step": 20862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.208576249011787,
+      "learning_rate": 1.2944127725733612e-09,
+      "loss": 0.9941,
+      "step": 20863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9484912505554517,
+      "learning_rate": 1.252321227868114e-09,
+      "loss": 1.1934,
+      "step": 20864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0295439377920514,
+      "learning_rate": 1.2109253881342764e-09,
+      "loss": 0.6379,
+      "step": 20865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.59258085264188,
+      "learning_rate": 1.1702252545237047e-09,
+      "loss": 0.9275,
+      "step": 20866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.970580329840066,
+      "learning_rate": 1.1302208281688265e-09,
+      "loss": 0.7034,
+      "step": 20867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7975751970017964,
+      "learning_rate": 1.0909121101826402e-09,
+      "loss": 0.2585,
+      "step": 20868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.569959714848387,
+      "learning_rate": 1.0522991016614914e-09,
+      "loss": 1.0684,
+      "step": 20869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5694451535427274,
+      "learning_rate": 1.0143818036739693e-09,
+      "loss": 1.2539,
+      "step": 20870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7711470481322036,
+      "learning_rate": 9.771602172803374e-10,
+      "loss": 1.1641,
+      "step": 20871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.249724265263269,
+      "learning_rate": 9.406343435158782e-10,
+      "loss": 0.5347,
+      "step": 20872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1005993305738775,
+      "learning_rate": 9.04804183396446e-10,
+      "loss": 0.772,
+      "step": 20873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2811009606574633,
+      "learning_rate": 8.696697379212415e-10,
+      "loss": 0.5328,
+      "step": 20874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8176220347691507,
+      "learning_rate": 8.352310080644854e-10,
+      "loss": 1.123,
+      "step": 20875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.796878893186983,
+      "learning_rate": 8.014879947837451e-10,
+      "loss": 1.5,
+      "step": 20876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.403404236566981,
+      "learning_rate": 7.6844069902271e-10,
+      "loss": 0.8682,
+      "step": 20877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7046693567340947,
+      "learning_rate": 7.360891216973142e-10,
+      "loss": 0.9248,
+      "step": 20878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.989094359852807,
+      "learning_rate": 7.044332637123897e-10,
+      "loss": 1.2598,
+      "step": 20879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3726246650510268,
+      "learning_rate": 6.734731259422367e-10,
+      "loss": 0.7563,
+      "step": 20880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.433433212272728,
+      "learning_rate": 6.432087092528294e-10,
+      "loss": 2.5488,
+      "step": 20881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.233250010286726,
+      "learning_rate": 6.136400144879373e-10,
+      "loss": 0.8184,
+      "step": 20882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9267868295022632,
+      "learning_rate": 5.847670424663498e-10,
+      "loss": 1.1484,
+      "step": 20883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.679833411487779,
+      "learning_rate": 5.565897939929787e-10,
+      "loss": 1.1699,
+      "step": 20884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7285390044102147,
+      "learning_rate": 5.291082698533067e-10,
+      "loss": 0.6816,
+      "step": 20885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.189530520846867,
+      "learning_rate": 5.023224708133878e-10,
+      "loss": 1.0977,
+      "step": 20886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0963145714172677,
+      "learning_rate": 4.762323976142957e-10,
+      "loss": 0.7065,
+      "step": 20887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.435371425691189,
+      "learning_rate": 4.5083805098322665e-10,
+      "loss": 1.082,
+      "step": 20888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.29713315992542,
+      "learning_rate": 4.2613943163072323e-10,
+      "loss": 1.4453,
+      "step": 20889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.561537608469033,
+      "learning_rate": 4.021365402395727e-10,
+      "loss": 1.3398,
+      "step": 20890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3500880691583226,
+      "learning_rate": 3.7882937747868443e-10,
+      "loss": 0.7109,
+      "step": 20891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.119442996165059,
+      "learning_rate": 3.5621794400031437e-10,
+      "loss": 0.5332,
+      "step": 20892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.507045330615042,
+      "learning_rate": 3.3430224042896307e-10,
+      "loss": 1.6074,
+      "step": 20893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.394723914158117,
+      "learning_rate": 3.130822673780287e-10,
+      "loss": 1.6025,
+      "step": 20894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.692374581320824,
+      "learning_rate": 2.9255802543592947e-10,
+      "loss": 0.9746,
+      "step": 20895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.449106995681068,
+      "learning_rate": 2.727295151744302e-10,
+      "loss": 0.8125,
+      "step": 20896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4555259401443155,
+      "learning_rate": 2.5359673714586695e-10,
+      "loss": 1.4912,
+      "step": 20897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.970811140464405,
+      "learning_rate": 2.351596918803711e-10,
+      "loss": 0.7705,
+      "step": 20898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.825390270992338,
+      "learning_rate": 2.1741837989419644e-10,
+      "loss": 0.6882,
+      "step": 20899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5957139992536544,
+      "learning_rate": 2.0037280168139216e-10,
+      "loss": 0.5186,
+      "step": 20900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9051604987789843,
+      "learning_rate": 1.8402295771380307e-10,
+      "loss": 1.1152,
+      "step": 20901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.017195932386133,
+      "learning_rate": 1.683688484466206e-10,
+      "loss": 1.5693,
+      "step": 20902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8973110294858464,
+      "learning_rate": 1.5341047431838284e-10,
+      "loss": 0.5364,
+      "step": 20903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5078632670460186,
+      "learning_rate": 1.3914783573987232e-10,
+      "loss": 1.5254,
+      "step": 20904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2310264411292993,
+      "learning_rate": 1.255809331135449e-10,
+      "loss": 1.2188,
+      "step": 20905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4314597032735836,
+      "learning_rate": 1.127097668168764e-10,
+      "loss": 0.6987,
+      "step": 20906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.455509654671423,
+      "learning_rate": 1.005343372023626e-10,
+      "loss": 0.9082,
+      "step": 20907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1579495794577057,
+      "learning_rate": 8.905464461417267e-11,
+      "loss": 0.7705,
+      "step": 20908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6589154501990793,
+      "learning_rate": 7.827068937149573e-11,
+      "loss": 1.0127,
+      "step": 20909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1277151152788027,
+      "learning_rate": 6.818247177131643e-11,
+      "loss": 0.8086,
+      "step": 20910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5373607864215644,
+      "learning_rate": 5.878999209674164e-11,
+      "loss": 1.0527,
+      "step": 20911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6167338006674008,
+      "learning_rate": 5.00932506058982e-11,
+      "loss": 0.6851,
+      "step": 20912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3400714640481466,
+      "learning_rate": 4.209224754581076e-11,
+      "loss": 1.291,
+      "step": 20913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1048998041618474,
+      "learning_rate": 3.478698313852391e-11,
+      "loss": 0.7129,
+      "step": 20914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.960357677812711,
+      "learning_rate": 2.8177457583877797e-11,
+      "loss": 1.0225,
+      "step": 20915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4757739234592013,
+      "learning_rate": 2.226367106783478e-11,
+      "loss": 0.9707,
+      "step": 20916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.695166925652677,
+      "learning_rate": 1.704562375415275e-11,
+      "loss": 1.4033,
+      "step": 20917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.48680560817133,
+      "learning_rate": 1.2523315787160705e-11,
+      "loss": 0.9688,
+      "step": 20918
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7845813985928844,
+      "learning_rate": 8.696747297309848e-12,
+      "loss": 0.9365,
+      "step": 20919
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9727240662614403,
+      "learning_rate": 5.56591838729581e-12,
+      "loss": 0.7314,
+      "step": 20920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.381876677305081,
+      "learning_rate": 3.1308291431608737e-12,
+      "loss": 1.5361,
+      "step": 20921
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.744961927801851,
+      "learning_rate": 1.3914796342939795e-12,
+      "loss": 1.1211,
+      "step": 20922
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9519552507704567,
+      "learning_rate": 3.4786991065516303e-13,
+      "loss": 0.8779,
+      "step": 20923
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.681520777461485,
+      "learning_rate": 0.0,
+      "loss": 1.1309,
+      "step": 20924
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.0417817831039429,
+      "eval_runtime": 884.1729,
+      "eval_samples_per_second": 15.777,
+      "eval_steps_per_second": 7.889,
+      "step": 20924
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20924,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.1620864978583552e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}