diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22442 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 1600,
+  "global_step": 3199,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00031259768677711783,
+      "grad_norm": 0.21875,
+      "learning_rate": 2e-05,
+      "loss": 2.2083,
+      "step": 1
+    },
+    {
+      "epoch": 0.00031259768677711783,
+      "eval_loss": 2.0116653442382812,
+      "eval_runtime": 1898.4475,
+      "eval_samples_per_second": 4.813,
+      "eval_steps_per_second": 2.407,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006251953735542357,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4e-05,
+      "loss": 2.0017,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009377930603313535,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6e-05,
+      "loss": 2.2668,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012503907471084713,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8e-05,
+      "loss": 1.9291,
+      "step": 4
+    },
+    {
+      "epoch": 0.0015629884338855893,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001,
+      "loss": 1.8984,
+      "step": 5
+    },
+    {
+      "epoch": 0.001875586120662707,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012,
+      "loss": 2.1924,
+      "step": 6
+    },
+    {
+      "epoch": 0.002188183807439825,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00014,
+      "loss": 2.135,
+      "step": 7
+    },
+    {
+      "epoch": 0.0025007814942169426,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00016,
+      "loss": 2.0434,
+      "step": 8
+    },
+    {
+      "epoch": 0.002813379180994061,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00018,
+      "loss": 1.892,
+      "step": 9
+    },
+    {
+      "epoch": 0.0031259768677711786,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002,
+      "loss": 2.0976,
+      "step": 10
+    },
+    {
+      "epoch": 0.0034385745545482964,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001999999969814363,
+      "loss": 2.257,
+      "step": 11
+    },
+    {
+      "epoch": 0.003751172241325414,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019999998792574533,
+      "loss": 2.0452,
+      "step": 12
+    },
+    {
+      "epoch": 0.004063769928102532,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019999997283292765,
+      "loss": 1.9728,
+      "step": 13
+    },
+    {
+      "epoch": 0.00437636761487965,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001999999517029842,
+      "loss": 2.0065,
+      "step": 14
+    },
+    {
+      "epoch": 0.004688965301656768,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019999992453591622,
+      "loss": 2.0628,
+      "step": 15
+    },
+    {
+      "epoch": 0.005001562988433885,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019999989133172538,
+      "loss": 2.0292,
+      "step": 16
+    },
+    {
+      "epoch": 0.0053141606752110035,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999985209041366,
+      "loss": 1.9125,
+      "step": 17
+    },
+    {
+      "epoch": 0.005626758361988122,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019999980681198345,
+      "loss": 1.8677,
+      "step": 18
+    },
+    {
+      "epoch": 0.005939356048765239,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019999975549643746,
+      "loss": 1.984,
+      "step": 19
+    },
+    {
+      "epoch": 0.006251953735542357,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019999969814377878,
+      "loss": 1.9886,
+      "step": 20
+    },
+    {
+      "epoch": 0.006564551422319475,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001999996347540109,
+      "loss": 2.0277,
+      "step": 21
+    },
+    {
+      "epoch": 0.006877149109096593,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001999995653271377,
+      "loss": 2.0216,
+      "step": 22
+    },
+    {
+      "epoch": 0.00718974679587371,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019999948986316324,
+      "loss": 1.8909,
+      "step": 23
+    },
+    {
+      "epoch": 0.007502344482650828,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019999940836209215,
+      "loss": 1.9551,
+      "step": 24
+    },
+    {
+      "epoch": 0.007814942169427946,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019999932082392937,
+      "loss": 1.9186,
+      "step": 25
+    },
+    {
+      "epoch": 0.008127539856205065,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019999922724868015,
+      "loss": 2.076,
+      "step": 26
+    },
+    {
+      "epoch": 0.008440137542982182,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019999912763635016,
+      "loss": 1.9682,
+      "step": 27
+    },
+    {
+      "epoch": 0.0087527352297593,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019999902198694543,
+      "loss": 1.8056,
+      "step": 28
+    },
+    {
+      "epoch": 0.009065332916536417,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019999891030047227,
+      "loss": 1.8404,
+      "step": 29
+    },
+    {
+      "epoch": 0.009377930603313536,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001999987925769375,
+      "loss": 2.113,
+      "step": 30
+    },
+    {
+      "epoch": 0.009690528290090653,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019999866881634815,
+      "loss": 1.8715,
+      "step": 31
+    },
+    {
+      "epoch": 0.01000312597686777,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00019999853901871175,
+      "loss": 3.09,
+      "step": 32
+    },
+    {
+      "epoch": 0.01031572366364489,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019999840318403613,
+      "loss": 2.1366,
+      "step": 33
+    },
+    {
+      "epoch": 0.010628321350422007,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019999826131232947,
+      "loss": 1.9007,
+      "step": 34
+    },
+    {
+      "epoch": 0.010940919037199124,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019999811340360034,
+      "loss": 1.9831,
+      "step": 35
+    },
+    {
+      "epoch": 0.011253516723976243,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001999979594578577,
+      "loss": 1.9656,
+      "step": 36
+    },
+    {
+      "epoch": 0.01156611441075336,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001999977994751108,
+      "loss": 2.8559,
+      "step": 37
+    },
+    {
+      "epoch": 0.011878712097530478,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019999763345536934,
+      "loss": 2.0553,
+      "step": 38
+    },
+    {
+      "epoch": 0.012191309784307595,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001999974613986433,
+      "loss": 1.8912,
+      "step": 39
+    },
+    {
+      "epoch": 0.012503907471084715,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001999972833049431,
+      "loss": 1.8342,
+      "step": 40
+    },
+    {
+      "epoch": 0.012816505157861832,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999709917427946,
+      "loss": 1.9032,
+      "step": 41
+    },
+    {
+      "epoch": 0.01312910284463895,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019999690900666353,
+      "loss": 1.9228,
+      "step": 42
+    },
+    {
+      "epoch": 0.013441700531416068,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019999671280210676,
+      "loss": 1.9666,
+      "step": 43
+    },
+    {
+      "epoch": 0.013754298218193186,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019999651056062102,
+      "loss": 1.901,
+      "step": 44
+    },
+    {
+      "epoch": 0.014066895904970303,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019999630228221852,
+      "loss": 2.1406,
+      "step": 45
+    },
+    {
+      "epoch": 0.01437949359174742,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001999960879669118,
+      "loss": 1.9389,
+      "step": 46
+    },
+    {
+      "epoch": 0.01469209127852454,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019999586761471384,
+      "loss": 1.9625,
+      "step": 47
+    },
+    {
+      "epoch": 0.015004688965301657,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019999564122563795,
+      "loss": 2.0456,
+      "step": 48
+    },
+    {
+      "epoch": 0.015317286652078774,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019999540879969775,
+      "loss": 1.9955,
+      "step": 49
+    },
+    {
+      "epoch": 0.01562988433885589,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999517033690727,
+      "loss": 1.8969,
+      "step": 50
+    },
+    {
+      "epoch": 0.01594248202563301,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019999492583728097,
+      "loss": 1.8544,
+      "step": 51
+    },
+    {
+      "epoch": 0.01625507971241013,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019999467530083356,
+      "loss": 1.779,
+      "step": 52
+    },
+    {
+      "epoch": 0.016567677399187245,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019999441872758017,
+      "loss": 1.8127,
+      "step": 53
+    },
+    {
+      "epoch": 0.016880275085964364,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001999941561175363,
+      "loss": 2.0516,
+      "step": 54
+    },
+    {
+      "epoch": 0.017192872772741483,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001999938874707178,
+      "loss": 2.1641,
+      "step": 55
+    },
+    {
+      "epoch": 0.0175054704595186,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019999361278714092,
+      "loss": 2.0483,
+      "step": 56
+    },
+    {
+      "epoch": 0.017818068146295718,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019999333206682218,
+      "loss": 1.9931,
+      "step": 57
+    },
+    {
+      "epoch": 0.018130665833072834,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019999304530977856,
+      "loss": 1.9795,
+      "step": 58
+    },
+    {
+      "epoch": 0.018443263519849953,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999275251602738,
+      "loss": 2.13,
+      "step": 59
+    },
+    {
+      "epoch": 0.018755861206627072,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001999924536855863,
+      "loss": 2.0652,
+      "step": 60
+    },
+    {
+      "epoch": 0.019068458893404187,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019999214881847338,
+      "loss": 1.9731,
+      "step": 61
+    },
+    {
+      "epoch": 0.019381056580181306,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019999183791470702,
+      "loss": 1.9303,
+      "step": 62
+    },
+    {
+      "epoch": 0.019693654266958426,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000199991520974306,
+      "loss": 1.9115,
+      "step": 63
+    },
+    {
+      "epoch": 0.02000625195373554,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001999911979972894,
+      "loss": 2.1622,
+      "step": 64
+    },
+    {
+      "epoch": 0.02031884964051266,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019999086898367678,
+      "loss": 1.9662,
+      "step": 65
+    },
+    {
+      "epoch": 0.02063144732728978,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019999053393348796,
+      "loss": 1.8382,
+      "step": 66
+    },
+    {
+      "epoch": 0.020944045014066895,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019999019284674317,
+      "loss": 1.9147,
+      "step": 67
+    },
+    {
+      "epoch": 0.021256642700844014,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019998984572346308,
+      "loss": 2.0712,
+      "step": 68
+    },
+    {
+      "epoch": 0.021569240387621133,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019998949256366854,
+      "loss": 2.0207,
+      "step": 69
+    },
+    {
+      "epoch": 0.02188183807439825,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019998913336738094,
+      "loss": 2.1334,
+      "step": 70
+    },
+    {
+      "epoch": 0.022194435761175368,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019998876813462192,
+      "loss": 2.7085,
+      "step": 71
+    },
+    {
+      "epoch": 0.022507033447952487,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019998839686541356,
+      "loss": 1.7364,
+      "step": 72
+    },
+    {
+      "epoch": 0.022819631134729602,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001999880195597783,
+      "loss": 1.9281,
+      "step": 73
+    },
+    {
+      "epoch": 0.02313222882150672,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019998763621773883,
+      "loss": 1.9648,
+      "step": 74
+    },
+    {
+      "epoch": 0.023444826508283837,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019998724683931838,
+      "loss": 1.9874,
+      "step": 75
+    },
+    {
+      "epoch": 0.023757424195060956,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001999868514245404,
+      "loss": 1.785,
+      "step": 76
+    },
+    {
+      "epoch": 0.024070021881838075,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001999864499734288,
+      "loss": 1.9094,
+      "step": 77
+    },
+    {
+      "epoch": 0.02438261956861519,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019998604248600777,
+      "loss": 1.9723,
+      "step": 78
+    },
+    {
+      "epoch": 0.02469521725539231,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019998562896230196,
+      "loss": 1.8739,
+      "step": 79
+    },
+    {
+      "epoch": 0.02500781494216943,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019998520940233636,
+      "loss": 1.936,
+      "step": 80
+    },
+    {
+      "epoch": 0.025320412628946545,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001999847838061362,
+      "loss": 1.8807,
+      "step": 81
+    },
+    {
+      "epoch": 0.025633010315723664,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019998435217372728,
+      "loss": 1.7412,
+      "step": 82
+    },
+    {
+      "epoch": 0.025945608002500783,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019998391450513556,
+      "loss": 1.8404,
+      "step": 83
+    },
+    {
+      "epoch": 0.0262582056892779,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019998347080038754,
+      "loss": 1.8108,
+      "step": 84
+    },
+    {
+      "epoch": 0.026570803376055017,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019998302105950994,
+      "loss": 2.0934,
+      "step": 85
+    },
+    {
+      "epoch": 0.026883401062832137,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019998256528252998,
+      "loss": 2.0021,
+      "step": 86
+    },
+    {
+      "epoch": 0.027195998749609252,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019998210346947515,
+      "loss": 1.9675,
+      "step": 87
+    },
+    {
+      "epoch": 0.02750859643638637,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019998163562037332,
+      "loss": 1.8488,
+      "step": 88
+    },
+    {
+      "epoch": 0.02782119412316349,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019998116173525272,
+      "loss": 1.9255,
+      "step": 89
+    },
+    {
+      "epoch": 0.028133791809940606,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000199980681814142,
+      "loss": 2.1055,
+      "step": 90
+    },
+    {
+      "epoch": 0.028446389496717725,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001999801958570701,
+      "loss": 2.1303,
+      "step": 91
+    },
+    {
+      "epoch": 0.02875898718349484,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019997970386406637,
+      "loss": 2.0517,
+      "step": 92
+    },
+    {
+      "epoch": 0.02907158487027196,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019997920583516053,
+      "loss": 1.8314,
+      "step": 93
+    },
+    {
+      "epoch": 0.02938418255704908,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001999787017703826,
+      "loss": 1.7953,
+      "step": 94
+    },
+    {
+      "epoch": 0.029696780243826194,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019997819166976308,
+      "loss": 1.8238,
+      "step": 95
+    },
+    {
+      "epoch": 0.030009377930603313,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001999776755333327,
+      "loss": 1.8905,
+      "step": 96
+    },
+    {
+      "epoch": 0.030321975617380433,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019997715336112263,
+      "loss": 1.7594,
+      "step": 97
+    },
+    {
+      "epoch": 0.030634573304157548,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001999766251531644,
+      "loss": 1.9648,
+      "step": 98
+    },
+    {
+      "epoch": 0.030947170990934667,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019997609090948996,
+      "loss": 2.1577,
+      "step": 99
+    },
+    {
+      "epoch": 0.03125976867771178,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001999755506301315,
+      "loss": 2.0563,
+      "step": 100
+    },
+    {
+      "epoch": 0.0315723663644889,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001999750043151216,
+      "loss": 1.8857,
+      "step": 101
+    },
+    {
+      "epoch": 0.03188496405126602,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019997445196449337,
+      "loss": 1.8832,
+      "step": 102
+    },
+    {
+      "epoch": 0.03219756173804314,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019997389357828,
+      "loss": 2.0352,
+      "step": 103
+    },
+    {
+      "epoch": 0.03251015942482026,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019997332915651532,
+      "loss": 2.0126,
+      "step": 104
+    },
+    {
+      "epoch": 0.03282275711159737,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019997275869923335,
+      "loss": 2.0201,
+      "step": 105
+    },
+    {
+      "epoch": 0.03313535479837449,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019997218220646853,
+      "loss": 2.3295,
+      "step": 106
+    },
+    {
+      "epoch": 0.03344795248515161,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001999715996782557,
+      "loss": 1.8919,
+      "step": 107
+    },
+    {
+      "epoch": 0.03376055017192873,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019997101111462998,
+      "loss": 1.7797,
+      "step": 108
+    },
+    {
+      "epoch": 0.03407314785870585,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019997041651562695,
+      "loss": 1.8956,
+      "step": 109
+    },
+    {
+      "epoch": 0.03438574554548297,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019996981588128244,
+      "loss": 1.9683,
+      "step": 110
+    },
+    {
+      "epoch": 0.03469834323226008,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019996920921163278,
+      "loss": 1.727,
+      "step": 111
+    },
+    {
+      "epoch": 0.0350109409190372,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019996859650671457,
+      "loss": 1.8966,
+      "step": 112
+    },
+    {
+      "epoch": 0.03532353860581432,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001999679777665648,
+      "loss": 1.8195,
+      "step": 113
+    },
+    {
+      "epoch": 0.035636136292591436,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001999673529912208,
+      "loss": 1.8785,
+      "step": 114
+    },
+    {
+      "epoch": 0.035948733979368555,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001999667221807203,
+      "loss": 1.9363,
+      "step": 115
+    },
+    {
+      "epoch": 0.03626133166614567,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019996608533510144,
+      "loss": 2.0314,
+      "step": 116
+    },
+    {
+      "epoch": 0.036573929352922786,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001999654424544026,
+      "loss": 1.9304,
+      "step": 117
+    },
+    {
+      "epoch": 0.036886527039699905,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001999647935386626,
+      "loss": 2.1456,
+      "step": 118
+    },
+    {
+      "epoch": 0.037199124726477024,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001999641385879206,
+      "loss": 1.6728,
+      "step": 119
+    },
+    {
+      "epoch": 0.037511722413254144,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019996347760221624,
+      "loss": 1.8201,
+      "step": 120
+    },
+    {
+      "epoch": 0.03782432010003126,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001999628105815893,
+      "loss": 2.096,
+      "step": 121
+    },
+    {
+      "epoch": 0.038136917786808375,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001999621375260801,
+      "loss": 1.9948,
+      "step": 122
+    },
+    {
+      "epoch": 0.038449515473585494,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001999614584357293,
+      "loss": 1.9562,
+      "step": 123
+    },
+    {
+      "epoch": 0.03876211316036261,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019996077331057788,
+      "loss": 1.8452,
+      "step": 124
+    },
+    {
+      "epoch": 0.03907471084713973,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019996008215066716,
+      "loss": 1.7615,
+      "step": 125
+    },
+    {
+      "epoch": 0.03938730853391685,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019995938495603893,
+      "loss": 1.7628,
+      "step": 126
+    },
+    {
+      "epoch": 0.03969990622069397,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019995868172673523,
+      "loss": 2.0241,
+      "step": 127
+    },
+    {
+      "epoch": 0.04001250390747108,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019995797246279856,
+      "loss": 2.0807,
+      "step": 128
+    },
+    {
+      "epoch": 0.0403251015942482,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019995725716427169,
+      "loss": 1.8564,
+      "step": 129
+    },
+    {
+      "epoch": 0.04063769928102532,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019995653583119785,
+      "loss": 2.1278,
+      "step": 130
+    },
+    {
+      "epoch": 0.04095029696780244,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019995580846362055,
+      "loss": 2.095,
+      "step": 131
+    },
+    {
+      "epoch": 0.04126289465457956,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019995507506158372,
+      "loss": 1.6848,
+      "step": 132
+    },
+    {
+      "epoch": 0.04157549234135667,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019995433562513163,
+      "loss": 1.8979,
+      "step": 133
+    },
+    {
+      "epoch": 0.04188809002813379,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019995359015430894,
+      "loss": 2.9492,
+      "step": 134
+    },
+    {
+      "epoch": 0.04220068771491091,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001999528386491606,
+      "loss": 1.8951,
+      "step": 135
+    },
+    {
+      "epoch": 0.04251328540168803,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019995208110973206,
+      "loss": 1.7656,
+      "step": 136
+    },
+    {
+      "epoch": 0.04282588308846515,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019995131753606902,
+      "loss": 2.0607,
+      "step": 137
+    },
+    {
+      "epoch": 0.043138480775242266,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019995054792821754,
+      "loss": 1.6803,
+      "step": 138
+    },
+    {
+      "epoch": 0.04345107846201938,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019994977228622414,
+      "loss": 2.0165,
+      "step": 139
+    },
+    {
+      "epoch": 0.0437636761487965,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001999489906101356,
+      "loss": 1.9388,
+      "step": 140
+    },
+    {
+      "epoch": 0.044076273835573616,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019994820289999913,
+      "loss": 1.6209,
+      "step": 141
+    },
+    {
+      "epoch": 0.044388871522350735,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001999474091558623,
+      "loss": 1.8157,
+      "step": 142
+    },
+    {
+      "epoch": 0.044701469209127855,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019994660937777301,
+      "loss": 1.7581,
+      "step": 143
+    },
+    {
+      "epoch": 0.045014066895904974,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00019994580356577957,
+      "loss": 2.6888,
+      "step": 144
+    },
+    {
+      "epoch": 0.045326664582682086,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019994499171993056,
+      "loss": 2.0103,
+      "step": 145
+    },
+    {
+      "epoch": 0.045639262269459205,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019994417384027507,
+      "loss": 1.7455,
+      "step": 146
+    },
+    {
+      "epoch": 0.045951859956236324,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019994334992686245,
+      "loss": 1.9287,
+      "step": 147
+    },
+    {
+      "epoch": 0.04626445764301344,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019994251997974241,
+      "loss": 1.8521,
+      "step": 148
+    },
+    {
+      "epoch": 0.04657705532979056,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019994168399896508,
+      "loss": 2.0915,
+      "step": 149
+    },
+    {
+      "epoch": 0.046889653016567674,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019994084198458097,
+      "loss": 2.0972,
+      "step": 150
+    },
+    {
+      "epoch": 0.04720225070334479,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019993999393664083,
+      "loss": 2.2031,
+      "step": 151
+    },
+    {
+      "epoch": 0.04751484839012191,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019993913985519592,
+      "loss": 1.8532,
+      "step": 152
+    },
+    {
+      "epoch": 0.04782744607689903,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001999382797402978,
+      "loss": 1.9278,
+      "step": 153
+    },
+    {
+      "epoch": 0.04814004376367615,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019993741359199834,
+      "loss": 1.6459,
+      "step": 154
+    },
+    {
+      "epoch": 0.04845264145045327,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001999365414103499,
+      "loss": 1.8459,
+      "step": 155
+    },
+    {
+      "epoch": 0.04876523913723038,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001999356631954051,
+      "loss": 1.9833,
+      "step": 156
+    },
+    {
+      "epoch": 0.0490778368240075,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019993477894721698,
+      "loss": 1.8361,
+      "step": 157
+    },
+    {
+      "epoch": 0.04939043451078462,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001999338886658389,
+      "loss": 2.0878,
+      "step": 158
+    },
+    {
+      "epoch": 0.04970303219756174,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001999329923513246,
+      "loss": 1.9454,
+      "step": 159
+    },
+    {
+      "epoch": 0.05001562988433886,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019993209000372818,
+      "loss": 1.982,
+      "step": 160
+    },
+    {
+      "epoch": 0.05032822757111598,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019993118162310415,
+      "loss": 1.9192,
+      "step": 161
+    },
+    {
+      "epoch": 0.05064082525789309,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001999302672095074,
+      "loss": 1.8865,
+      "step": 162
+    },
+    {
+      "epoch": 0.05095342294467021,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019992934676299302,
+      "loss": 1.6733,
+      "step": 163
+    },
+    {
+      "epoch": 0.05126602063144733,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019992842028361665,
+      "loss": 1.9374,
+      "step": 164
+    },
+    {
+      "epoch": 0.051578618318224446,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001999274877714342,
+      "loss": 1.9537,
+      "step": 165
+    },
+    {
+      "epoch": 0.051891216005001566,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000199926549226502,
+      "loss": 1.8767,
+      "step": 166
+    },
+    {
+      "epoch": 0.05220381369177868,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019992560464887667,
+      "loss": 1.8994,
+      "step": 167
+    },
+    {
+      "epoch": 0.0525164113785558,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019992465403861524,
+      "loss": 1.7415,
+      "step": 168
+    },
+    {
+      "epoch": 0.052829009065332916,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019992369739577512,
+      "loss": 1.7688,
+      "step": 169
+    },
+    {
+      "epoch": 0.053141606752110035,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019992273472041404,
+      "loss": 1.7507,
+      "step": 170
+    },
+    {
+      "epoch": 0.053454204438887154,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019992176601259015,
+      "loss": 1.995,
+      "step": 171
+    },
+    {
+      "epoch": 0.05376680212566427,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019992079127236192,
+      "loss": 1.9025,
+      "step": 172
+    },
+    {
+      "epoch": 0.054079399812441385,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001999198104997882,
+      "loss": 1.7634,
+      "step": 173
+    },
+    {
+      "epoch": 0.054391997499218504,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019991882369492815,
+      "loss": 1.8371,
+      "step": 174
+    },
+    {
+      "epoch": 0.05470459518599562,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001999178308578414,
+      "loss": 1.7978,
+      "step": 175
+    },
+    {
+      "epoch": 0.05501719287277274,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001999168319885879,
+      "loss": 2.0066,
+      "step": 176
+    },
+    {
+      "epoch": 0.05532979055954986,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019991582708722792,
+      "loss": 1.6957,
+      "step": 177
+    },
+    {
+      "epoch": 0.05564238824632698,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001999148161538221,
+      "loss": 1.8989,
+      "step": 178
+    },
+    {
+      "epoch": 0.05595498593310409,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019991379918843155,
+      "loss": 2.0687,
+      "step": 179
+    },
+    {
+      "epoch": 0.05626758361988121,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019991277619111763,
+      "loss": 1.9398,
+      "step": 180
+    },
+    {
+      "epoch": 0.05658018130665833,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019991174716194203,
+      "loss": 1.7309,
+      "step": 181
+    },
+    {
+      "epoch": 0.05689277899343545,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019991071210096698,
+      "loss": 1.8865,
+      "step": 182
+    },
+    {
+      "epoch": 0.05720537668021257,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019990967100825491,
+      "loss": 1.8802,
+      "step": 183
+    },
+    {
+      "epoch": 0.05751797436698968,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001999086238838687,
+      "loss": 1.814,
+      "step": 184
+    },
+    {
+      "epoch": 0.0578305720537668,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019990757072787152,
+      "loss": 1.6507,
+      "step": 185
+    },
+    {
+      "epoch": 0.05814316974054392,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000199906511540327,
+      "loss": 1.921,
+      "step": 186
+    },
+    {
+      "epoch": 0.05845576742732104,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001999054463212991,
+      "loss": 1.9151,
+      "step": 187
+    },
+    {
+      "epoch": 0.05876836511409816,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019990437507085202,
+      "loss": 2.0727,
+      "step": 188
+    },
+    {
+      "epoch": 0.05908096280087528,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019990329778905058,
+      "loss": 2.0359,
+      "step": 189
+    },
+    {
+      "epoch": 0.05939356048765239,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019990221447595968,
+      "loss": 1.9311,
+      "step": 190
+    },
+    {
+      "epoch": 0.05970615817442951,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019990112513164484,
+      "loss": 1.8018,
+      "step": 191
+    },
+    {
+      "epoch": 0.06001875586120663,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019990002975617174,
+      "loss": 1.9104,
+      "step": 192
+    },
+    {
+      "epoch": 0.060331353547983746,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019989892834960656,
+      "loss": 1.7227,
+      "step": 193
+    },
+    {
+      "epoch": 0.060643951234760865,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019989782091201573,
+      "loss": 1.7287,
+      "step": 194
+    },
+    {
+      "epoch": 0.06095654892153798,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001998967074434662,
+      "loss": 1.8525,
+      "step": 195
+    },
+    {
+      "epoch": 0.061269146608315096,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019989558794402515,
+      "loss": 2.4259,
+      "step": 196
+    },
+    {
+      "epoch": 0.061581744295092215,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001998944624137601,
+      "loss": 2.1134,
+      "step": 197
+    },
+    {
+      "epoch": 0.061894341981869334,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001998933308527391,
+      "loss": 1.9239,
+      "step": 198
+    },
+    {
+      "epoch": 0.06220693966864645,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001998921932610304,
+      "loss": 1.7292,
+      "step": 199
+    },
+    {
+      "epoch": 0.06251953735542357,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001998910496387027,
+      "loss": 1.7629,
+      "step": 200
+    },
+    {
+      "epoch": 0.06283213504220068,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019988989998582506,
+      "loss": 2.005,
+      "step": 201
+    },
+    {
+      "epoch": 0.0631447327289778,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019988874430246686,
+      "loss": 1.7605,
+      "step": 202
+    },
+    {
+      "epoch": 0.06345733041575492,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001998875825886979,
+      "loss": 1.748,
+      "step": 203
+    },
+    {
+      "epoch": 0.06376992810253204,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019988641484458826,
+      "loss": 2.1037,
+      "step": 204
+    },
+    {
+      "epoch": 0.06408252578930916,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019988524107020846,
+      "loss": 1.9274,
+      "step": 205
+    },
+    {
+      "epoch": 0.06439512347608628,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019988406126562937,
+      "loss": 1.7823,
+      "step": 206
+    },
+    {
+      "epoch": 0.0647077211628634,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019988287543092225,
+      "loss": 2.06,
+      "step": 207
+    },
+    {
+      "epoch": 0.06502031884964052,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019988168356615865,
+      "loss": 1.9327,
+      "step": 208
+    },
+    {
+      "epoch": 0.06533291653641764,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019988048567141052,
+      "loss": 1.9889,
+      "step": 209
+    },
+    {
+      "epoch": 0.06564551422319474,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019987928174675023,
+      "loss": 1.6262,
+      "step": 210
+    },
+    {
+      "epoch": 0.06595811190997186,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019987807179225035,
+      "loss": 1.8805,
+      "step": 211
+    },
+    {
+      "epoch": 0.06627070959674898,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019987685580798403,
+      "loss": 1.7265,
+      "step": 212
+    },
+    {
+      "epoch": 0.0665833072835261,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001998756337940247,
+      "loss": 1.7049,
+      "step": 213
+    },
+    {
+      "epoch": 0.06689590497030322,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019987440575044602,
+      "loss": 1.7256,
+      "step": 214
+    },
+    {
+      "epoch": 0.06720850265708034,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019987317167732222,
+      "loss": 1.9469,
+      "step": 215
+    },
+    {
+      "epoch": 0.06752110034385746,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019987193157472777,
+      "loss": 2.0254,
+      "step": 216
+    },
+    {
+      "epoch": 0.06783369803063458,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019987068544273756,
+      "loss": 2.1006,
+      "step": 217
+    },
+    {
+      "epoch": 0.0681462957174117,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019986943328142678,
+      "loss": 1.9486,
+      "step": 218
+    },
+    {
+      "epoch": 0.06845889340418881,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019986817509087107,
+      "loss": 1.9707,
+      "step": 219
+    },
+    {
+      "epoch": 0.06877149109096593,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019986691087114635,
+      "loss": 1.868,
+      "step": 220
+    },
+    {
+      "epoch": 0.06908408877774304,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019986564062232897,
+      "loss": 1.9028,
+      "step": 221
+    },
+    {
+      "epoch": 0.06939668646452016,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001998643643444956,
+      "loss": 1.9136,
+      "step": 222
+    },
+    {
+      "epoch": 0.06970928415129728,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001998630820377233,
+      "loss": 1.8039,
+      "step": 223
+    },
+    {
+      "epoch": 0.0700218818380744,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019986179370208947,
+      "loss": 1.7326,
+      "step": 224
+    },
+    {
+      "epoch": 0.07033447952485151,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001998604993376719,
+      "loss": 1.7712,
+      "step": 225
+    },
+    {
+      "epoch": 0.07064707721162863,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019985919894454875,
+      "loss": 1.9061,
+      "step": 226
+    },
+    {
+      "epoch": 0.07095967489840575,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019985789252279846,
+      "loss": 1.8444,
+      "step": 227
+    },
+    {
+      "epoch": 0.07127227258518287,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001998565800725,
+      "loss": 2.1696,
+      "step": 228
+    },
+    {
+      "epoch": 0.07158487027195999,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019985526159373255,
+      "loss": 1.9888,
+      "step": 229
+    },
+    {
+      "epoch": 0.07189746795873711,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019985393708657568,
+      "loss": 2.018,
+      "step": 230
+    },
+    {
+      "epoch": 0.07221006564551423,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001998526065511094,
+      "loss": 1.7847,
+      "step": 231
+    },
+    {
+      "epoch": 0.07252266333229133,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019985126998741404,
+      "loss": 1.879,
+      "step": 232
+    },
+    {
+      "epoch": 0.07283526101906845,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019984992739557024,
+      "loss": 1.7065,
+      "step": 233
+    },
+    {
+      "epoch": 0.07314785870584557,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019984857877565907,
+      "loss": 1.7451,
+      "step": 234
+    },
+    {
+      "epoch": 0.07346045639262269,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.000199847224127762,
+      "loss": 1.8228,
+      "step": 235
+    },
+    {
+      "epoch": 0.07377305407939981,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019984586345196074,
+      "loss": 1.9904,
+      "step": 236
+    },
+    {
+      "epoch": 0.07408565176617693,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001998444967483375,
+      "loss": 1.8958,
+      "step": 237
+    },
+    {
+      "epoch": 0.07439824945295405,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019984312401697473,
+      "loss": 1.8913,
+      "step": 238
+    },
+    {
+      "epoch": 0.07471084713973117,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019984174525795536,
+      "loss": 1.9273,
+      "step": 239
+    },
+    {
+      "epoch": 0.07502344482650829,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019984036047136257,
+      "loss": 1.8831,
+      "step": 240
+    },
+    {
+      "epoch": 0.0753360425132854,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019983896965728001,
+      "loss": 1.9506,
+      "step": 241
+    },
+    {
+      "epoch": 0.07564864020006253,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019983757281579162,
+      "loss": 1.971,
+      "step": 242
+    },
+    {
+      "epoch": 0.07596123788683964,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019983616994698173,
+      "loss": 1.8156,
+      "step": 243
+    },
+    {
+      "epoch": 0.07627383557361675,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019983476105093505,
+      "loss": 1.9397,
+      "step": 244
+    },
+    {
+      "epoch": 0.07658643326039387,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019983334612773662,
+      "loss": 1.7567,
+      "step": 245
+    },
+    {
+      "epoch": 0.07689903094717099,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019983192517747186,
+      "loss": 1.8685,
+      "step": 246
+    },
+    {
+      "epoch": 0.0772116286339481,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019983049820022656,
+      "loss": 2.2285,
+      "step": 247
+    },
+    {
+      "epoch": 0.07752422632072523,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019982906519608687,
+      "loss": 1.9532,
+      "step": 248
+    },
+    {
+      "epoch": 0.07783682400750234,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001998276261651393,
+      "loss": 1.8775,
+      "step": 249
+    },
+    {
+      "epoch": 0.07814942169427946,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019982618110747074,
+      "loss": 1.892,
+      "step": 250
+    },
+    {
+      "epoch": 0.07846201938105658,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019982473002316838,
+      "loss": 2.2827,
+      "step": 251
+    },
+    {
+      "epoch": 0.0787746170678337,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001998232729123199,
+      "loss": 2.1452,
+      "step": 252
+    },
+    {
+      "epoch": 0.07908721475461082,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019982180977501322,
+      "loss": 1.7888,
+      "step": 253
+    },
+    {
+      "epoch": 0.07939981244138794,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019982034061133666,
+      "loss": 1.7486,
+      "step": 254
+    },
+    {
+      "epoch": 0.07971241012816505,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019981886542137892,
+      "loss": 1.8143,
+      "step": 255
+    },
+    {
+      "epoch": 0.08002500781494216,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019981738420522913,
+      "loss": 1.839,
+      "step": 256
+    },
+    {
+      "epoch": 0.08033760550171928,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019981589696297663,
+      "loss": 1.918,
+      "step": 257
+    },
+    {
+      "epoch": 0.0806502031884964,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019981440369471124,
+      "loss": 1.9144,
+      "step": 258
+    },
+    {
+      "epoch": 0.08096280087527352,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019981290440052306,
+      "loss": 1.7846,
+      "step": 259
+    },
+    {
+      "epoch": 0.08127539856205064,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001998113990805027,
+      "loss": 1.9837,
+      "step": 260
+    },
+    {
+      "epoch": 0.08158799624882776,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019980988773474098,
+      "loss": 1.9422,
+      "step": 261
+    },
+    {
+      "epoch": 0.08190059393560488,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019980837036332917,
+      "loss": 1.7637,
+      "step": 262
+    },
+    {
+      "epoch": 0.082213191622382,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001998068469663588,
+      "loss": 2.5924,
+      "step": 263
+    },
+    {
+      "epoch": 0.08252578930915912,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001998053175439219,
+      "loss": 1.8041,
+      "step": 264
+    },
+    {
+      "epoch": 0.08283838699593624,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019980378209611083,
+      "loss": 2.139,
+      "step": 265
+    },
+    {
+      "epoch": 0.08315098468271334,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001998022406230182,
+      "loss": 1.8233,
+      "step": 266
+    },
+    {
+      "epoch": 0.08346358236949046,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001998006931247372,
+      "loss": 1.9227,
+      "step": 267
+    },
+    {
+      "epoch": 0.08377618005626758,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019979913960136114,
+      "loss": 1.7389,
+      "step": 268
+    },
+    {
+      "epoch": 0.0840887777430447,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019979758005298385,
+      "loss": 1.6342,
+      "step": 269
+    },
+    {
+      "epoch": 0.08440137542982182,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001997960144796995,
+      "loss": 1.9472,
+      "step": 270
+    },
+    {
+      "epoch": 0.08471397311659894,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019979444288160253,
+      "loss": 1.7985,
+      "step": 271
+    },
+    {
+      "epoch": 0.08502657080337606,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019979286525878792,
+      "loss": 1.8546,
+      "step": 272
+    },
+    {
+      "epoch": 0.08533916849015317,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019979128161135083,
+      "loss": 1.9697,
+      "step": 273
+    },
+    {
+      "epoch": 0.0856517661769303,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019978969193938694,
+      "loss": 2.095,
+      "step": 274
+    },
+    {
+      "epoch": 0.08596436386370741,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019978809624299218,
+      "loss": 1.9491,
+      "step": 275
+    },
+    {
+      "epoch": 0.08627696155048453,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019978649452226285,
+      "loss": 1.9463,
+      "step": 276
+    },
+    {
+      "epoch": 0.08658955923726164,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019978488677729574,
+      "loss": 1.8981,
+      "step": 277
+    },
+    {
+      "epoch": 0.08690215692403876,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019978327300818784,
+      "loss": 1.9126,
+      "step": 278
+    },
+    {
+      "epoch": 0.08721475461081588,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001997816532150366,
+      "loss": 1.8987,
+      "step": 279
+    },
+    {
+      "epoch": 0.087527352297593,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019978002739793978,
+      "loss": 1.7486,
+      "step": 280
+    },
+    {
+      "epoch": 0.08783994998437011,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019977839555699553,
+      "loss": 1.9603,
+      "step": 281
+    },
+    {
+      "epoch": 0.08815254767114723,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019977675769230246,
+      "loss": 1.8714,
+      "step": 282
+    },
+    {
+      "epoch": 0.08846514535792435,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019977511380395933,
+      "loss": 2.0087,
+      "step": 283
+    },
+    {
+      "epoch": 0.08877774304470147,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019977346389206545,
+      "loss": 2.1653,
+      "step": 284
+    },
+    {
+      "epoch": 0.08909034073147859,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019977180795672044,
+      "loss": 2.0311,
+      "step": 285
+    },
+    {
+      "epoch": 0.08940293841825571,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019977014599802418,
+      "loss": 1.8212,
+      "step": 286
+    },
+    {
+      "epoch": 0.08971553610503283,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019976847801607712,
+      "loss": 2.0245,
+      "step": 287
+    },
+    {
+      "epoch": 0.09002813379180995,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001997668040109799,
+      "loss": 1.8573,
+      "step": 288
+    },
+    {
+      "epoch": 0.09034073147858705,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019976512398283357,
+      "loss": 1.7208,
+      "step": 289
+    },
+    {
+      "epoch": 0.09065332916536417,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019976343793173958,
+      "loss": 1.7056,
+      "step": 290
+    },
+    {
+      "epoch": 0.09096592685214129,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019976174585779972,
+      "loss": 1.8874,
+      "step": 291
+    },
+    {
+      "epoch": 0.09127852453891841,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019976004776111613,
+      "loss": 1.5886,
+      "step": 292
+    },
+    {
+      "epoch": 0.09159112222569553,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019975834364179134,
+      "loss": 1.7725,
+      "step": 293
+    },
+    {
+      "epoch": 0.09190371991247265,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001997566334999282,
+      "loss": 1.7855,
+      "step": 294
+    },
+    {
+      "epoch": 0.09221631759924977,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019975491733563,
+      "loss": 1.7919,
+      "step": 295
+    },
+    {
+      "epoch": 0.09252891528602689,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019975319514900028,
+      "loss": 1.7353,
+      "step": 296
+    },
+    {
+      "epoch": 0.092841512972804,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019975146694014312,
+      "loss": 1.8983,
+      "step": 297
+    },
+    {
+      "epoch": 0.09315411065958112,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019974973270916273,
+      "loss": 2.115,
+      "step": 298
+    },
+    {
+      "epoch": 0.09346670834635824,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019974799245616387,
+      "loss": 1.9605,
+      "step": 299
+    },
+    {
+      "epoch": 0.09377930603313535,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001997462461812516,
+      "loss": 1.9963,
+      "step": 300
+    },
+    {
+      "epoch": 0.09409190371991247,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019974449388453135,
+      "loss": 1.8288,
+      "step": 301
+    },
+    {
+      "epoch": 0.09440450140668959,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001997427355661089,
+      "loss": 1.7948,
+      "step": 302
+    },
+    {
+      "epoch": 0.0947170990934667,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001997409712260904,
+      "loss": 1.868,
+      "step": 303
+    },
+    {
+      "epoch": 0.09502969678024382,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019973920086458237,
+      "loss": 1.8929,
+      "step": 304
+    },
+    {
+      "epoch": 0.09534229446702094,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019973742448169165,
+      "loss": 1.6884,
+      "step": 305
+    },
+    {
+      "epoch": 0.09565489215379806,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019973564207752554,
+      "loss": 1.6901,
+      "step": 306
+    },
+    {
+      "epoch": 0.09596748984057518,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019973385365219164,
+      "loss": 1.7943,
+      "step": 307
+    },
+    {
+      "epoch": 0.0962800875273523,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001997320592057979,
+      "loss": 1.9581,
+      "step": 308
+    },
+    {
+      "epoch": 0.09659268521412942,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019973025873845263,
+      "loss": 1.6522,
+      "step": 309
+    },
+    {
+      "epoch": 0.09690528290090654,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019972845225026456,
+      "loss": 1.9327,
+      "step": 310
+    },
+    {
+      "epoch": 0.09721788058768364,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019972663974134275,
+      "loss": 1.9542,
+      "step": 311
+    },
+    {
+      "epoch": 0.09753047827446076,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019972482121179664,
+      "loss": 2.0571,
+      "step": 312
+    },
+    {
+      "epoch": 0.09784307596123788,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019972299666173594,
+      "loss": 2.2707,
+      "step": 313
+    },
+    {
+      "epoch": 0.098155673648015,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001997211660912709,
+      "loss": 1.9587,
+      "step": 314
+    },
+    {
+      "epoch": 0.09846827133479212,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019971932950051198,
+      "loss": 2.0126,
+      "step": 315
+    },
+    {
+      "epoch": 0.09878086902156924,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019971748688957003,
+      "loss": 1.7935,
+      "step": 316
+    },
+    {
+      "epoch": 0.09909346670834636,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019971563825855638,
+      "loss": 1.8761,
+      "step": 317
+    },
+    {
+      "epoch": 0.09940606439512348,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019971378360758254,
+      "loss": 2.2404,
+      "step": 318
+    },
+    {
+      "epoch": 0.0997186620819006,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001997119229367605,
+      "loss": 1.8394,
+      "step": 319
+    },
+    {
+      "epoch": 0.10003125976867772,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019971005624620265,
+      "loss": 1.8923,
+      "step": 320
+    },
+    {
+      "epoch": 0.10034385745545484,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019970818353602163,
+      "loss": 1.6077,
+      "step": 321
+    },
+    {
+      "epoch": 0.10065645514223195,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019970630480633047,
+      "loss": 1.8617,
+      "step": 322
+    },
+    {
+      "epoch": 0.10096905282900906,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001997044200572427,
+      "loss": 1.892,
+      "step": 323
+    },
+    {
+      "epoch": 0.10128165051578618,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000199702529288872,
+      "loss": 1.7457,
+      "step": 324
+    },
+    {
+      "epoch": 0.1015942482025633,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019970063250133256,
+      "loss": 1.9309,
+      "step": 325
+    },
+    {
+      "epoch": 0.10190684588934042,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019969872969473888,
+      "loss": 1.905,
+      "step": 326
+    },
+    {
+      "epoch": 0.10221944357611754,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019969682086920585,
+      "loss": 1.697,
+      "step": 327
+    },
+    {
+      "epoch": 0.10253204126289465,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001996949060248487,
+      "loss": 1.8728,
+      "step": 328
+    },
+    {
+      "epoch": 0.10284463894967177,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019969298516178303,
+      "loss": 1.7783,
+      "step": 329
+    },
+    {
+      "epoch": 0.10315723663644889,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001996910582801248,
+      "loss": 1.8591,
+      "step": 330
+    },
+    {
+      "epoch": 0.10346983432322601,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019968912537999034,
+      "loss": 1.8009,
+      "step": 331
+    },
+    {
+      "epoch": 0.10378243201000313,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019968718646149635,
+      "loss": 1.6679,
+      "step": 332
+    },
+    {
+      "epoch": 0.10409502969678025,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019968524152475986,
+      "loss": 1.9598,
+      "step": 333
+    },
+    {
+      "epoch": 0.10440762738355736,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019968329056989836,
+      "loss": 1.7525,
+      "step": 334
+    },
+    {
+      "epoch": 0.10472022507033447,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019968133359702956,
+      "loss": 1.9891,
+      "step": 335
+    },
+    {
+      "epoch": 0.1050328227571116,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019967937060627163,
+      "loss": 2.6398,
+      "step": 336
+    },
+    {
+      "epoch": 0.10534542044388871,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019967740159774304,
+      "loss": 1.8126,
+      "step": 337
+    },
+    {
+      "epoch": 0.10565801813066583,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001996754265715627,
+      "loss": 1.5844,
+      "step": 338
+    },
+    {
+      "epoch": 0.10597061581744295,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019967344552784987,
+      "loss": 2.6948,
+      "step": 339
+    },
+    {
+      "epoch": 0.10628321350422007,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019967145846672412,
+      "loss": 1.8124,
+      "step": 340
+    },
+    {
+      "epoch": 0.10659581119099719,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019966946538830537,
+      "loss": 1.7512,
+      "step": 341
+    },
+    {
+      "epoch": 0.10690840887777431,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019966746629271402,
+      "loss": 1.886,
+      "step": 342
+    },
+    {
+      "epoch": 0.10722100656455143,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001996654611800707,
+      "loss": 1.8067,
+      "step": 343
+    },
+    {
+      "epoch": 0.10753360425132855,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001996634500504965,
+      "loss": 1.8013,
+      "step": 344
+    },
+    {
+      "epoch": 0.10784620193810565,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019966143290411282,
+      "loss": 1.701,
+      "step": 345
+    },
+    {
+      "epoch": 0.10815879962488277,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019965940974104145,
+      "loss": 1.6386,
+      "step": 346
+    },
+    {
+      "epoch": 0.10847139731165989,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001996573805614045,
+      "loss": 1.9652,
+      "step": 347
+    },
+    {
+      "epoch": 0.10878399499843701,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001996553453653245,
+      "loss": 1.8178,
+      "step": 348
+    },
+    {
+      "epoch": 0.10909659268521413,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019965330415292428,
+      "loss": 1.8802,
+      "step": 349
+    },
+    {
+      "epoch": 0.10940919037199125,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001996512569243271,
+      "loss": 1.6879,
+      "step": 350
+    },
+    {
+      "epoch": 0.10972178805876837,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001996492036796566,
+      "loss": 1.8288,
+      "step": 351
+    },
+    {
+      "epoch": 0.11003438574554548,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019964714441903663,
+      "loss": 1.8453,
+      "step": 352
+    },
+    {
+      "epoch": 0.1103469834323226,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019964507914259157,
+      "loss": 1.8259,
+      "step": 353
+    },
+    {
+      "epoch": 0.11065958111909972,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019964300785044615,
+      "loss": 1.9748,
+      "step": 354
+    },
+    {
+      "epoch": 0.11097217880587684,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019964093054272535,
+      "loss": 2.0296,
+      "step": 355
+    },
+    {
+      "epoch": 0.11128477649265396,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001996388472195546,
+      "loss": 2.1065,
+      "step": 356
+    },
+    {
+      "epoch": 0.11159737417943107,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019963675788105967,
+      "loss": 1.712,
+      "step": 357
+    },
+    {
+      "epoch": 0.11190997186620819,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001996346625273667,
+      "loss": 2.178,
+      "step": 358
+    },
+    {
+      "epoch": 0.1122225695529853,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019963256115860219,
+      "loss": 1.6854,
+      "step": 359
+    },
+    {
+      "epoch": 0.11253516723976242,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019963045377489297,
+      "loss": 1.7912,
+      "step": 360
+    },
+    {
+      "epoch": 0.11284776492653954,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019962834037636634,
+      "loss": 1.7385,
+      "step": 361
+    },
+    {
+      "epoch": 0.11316036261331666,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019962622096314983,
+      "loss": 1.787,
+      "step": 362
+    },
+    {
+      "epoch": 0.11347296030009378,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019962409553537141,
+      "loss": 1.7083,
+      "step": 363
+    },
+    {
+      "epoch": 0.1137855579868709,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019962196409315937,
+      "loss": 1.7489,
+      "step": 364
+    },
+    {
+      "epoch": 0.11409815567364802,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019961982663664244,
+      "loss": 1.8184,
+      "step": 365
+    },
+    {
+      "epoch": 0.11441075336042514,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001996176831659496,
+      "loss": 1.924,
+      "step": 366
+    },
+    {
+      "epoch": 0.11472335104720226,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001996155336812103,
+      "loss": 2.1837,
+      "step": 367
+    },
+    {
+      "epoch": 0.11503594873397936,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019961337818255424,
+      "loss": 1.9305,
+      "step": 368
+    },
+    {
+      "epoch": 0.11534854642075648,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019961121667011166,
+      "loss": 1.9867,
+      "step": 369
+    },
+    {
+      "epoch": 0.1156611441075336,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019960904914401298,
+      "loss": 1.968,
+      "step": 370
+    },
+    {
+      "epoch": 0.11597374179431072,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019960687560438908,
+      "loss": 1.6922,
+      "step": 371
+    },
+    {
+      "epoch": 0.11628633948108784,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019960469605137114,
+      "loss": 1.7978,
+      "step": 372
+    },
+    {
+      "epoch": 0.11659893716786496,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001996025104850908,
+      "loss": 1.8674,
+      "step": 373
+    },
+    {
+      "epoch": 0.11691153485464208,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019960031890567997,
+      "loss": 1.7445,
+      "step": 374
+    },
+    {
+      "epoch": 0.1172241325414192,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019959812131327095,
+      "loss": 1.7513,
+      "step": 375
+    },
+    {
+      "epoch": 0.11753673022819631,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019959591770799643,
+      "loss": 1.7463,
+      "step": 376
+    },
+    {
+      "epoch": 0.11784932791497343,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019959370808998945,
+      "loss": 1.6496,
+      "step": 377
+    },
+    {
+      "epoch": 0.11816192560175055,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001995914924593834,
+      "loss": 1.6407,
+      "step": 378
+    },
+    {
+      "epoch": 0.11847452328852766,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019958927081631205,
+      "loss": 1.9992,
+      "step": 379
+    },
+    {
+      "epoch": 0.11878712097530478,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001995870431609095,
+      "loss": 1.7538,
+      "step": 380
+    },
+    {
+      "epoch": 0.1190997186620819,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019958480949331024,
+      "loss": 1.6851,
+      "step": 381
+    },
+    {
+      "epoch": 0.11941231634885902,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019958256981364916,
+      "loss": 1.7887,
+      "step": 382
+    },
+    {
+      "epoch": 0.11972491403563613,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019958032412206142,
+      "loss": 1.8162,
+      "step": 383
+    },
+    {
+      "epoch": 0.12003751172241325,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001995780724186826,
+      "loss": 1.8541,
+      "step": 384
+    },
+    {
+      "epoch": 0.12035010940919037,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019957581470364869,
+      "loss": 1.8194,
+      "step": 385
+    },
+    {
+      "epoch": 0.12066270709596749,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001995735509770959,
+      "loss": 1.7891,
+      "step": 386
+    },
+    {
+      "epoch": 0.12097530478274461,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019957128123916103,
+      "loss": 1.992,
+      "step": 387
+    },
+    {
+      "epoch": 0.12128790246952173,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019956900548998097,
+      "loss": 1.9259,
+      "step": 388
+    },
+    {
+      "epoch": 0.12160050015629885,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019956672372969315,
+      "loss": 2.0642,
+      "step": 389
+    },
+    {
+      "epoch": 0.12191309784307595,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001995644359584354,
+      "loss": 1.6211,
+      "step": 390
+    },
+    {
+      "epoch": 0.12222569552985307,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019956214217634575,
+      "loss": 1.7604,
+      "step": 391
+    },
+    {
+      "epoch": 0.12253829321663019,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019955984238356268,
+      "loss": 1.8761,
+      "step": 392
+    },
+    {
+      "epoch": 0.12285089090340731,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001995575365802251,
+      "loss": 2.0069,
+      "step": 393
+    },
+    {
+      "epoch": 0.12316348859018443,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001995552247664721,
+      "loss": 1.7372,
+      "step": 394
+    },
+    {
+      "epoch": 0.12347608627696155,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019955290694244338,
+      "loss": 1.8025,
+      "step": 395
+    },
+    {
+      "epoch": 0.12378868396373867,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019955058310827878,
+      "loss": 1.8633,
+      "step": 396
+    },
+    {
+      "epoch": 0.12410128165051579,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019954825326411863,
+      "loss": 1.9765,
+      "step": 397
+    },
+    {
+      "epoch": 0.1244138793372929,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001995459174101036,
+      "loss": 1.6959,
+      "step": 398
+    },
+    {
+      "epoch": 0.12472647702407003,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001995435755463746,
+      "loss": 1.6401,
+      "step": 399
+    },
+    {
+      "epoch": 0.12503907471084713,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019954122767307318,
+      "loss": 2.1424,
+      "step": 400
+    },
+    {
+      "epoch": 0.12535167239762426,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019953887379034094,
+      "loss": 1.9393,
+      "step": 401
+    },
+    {
+      "epoch": 0.12566427008440137,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019953651389832008,
+      "loss": 1.8414,
+      "step": 402
+    },
+    {
+      "epoch": 0.1259768677711785,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019953414799715304,
+      "loss": 1.9348,
+      "step": 403
+    },
+    {
+      "epoch": 0.1262894654579556,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019953177608698263,
+      "loss": 1.6774,
+      "step": 404
+    },
+    {
+      "epoch": 0.12660206314473274,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019952939816795205,
+      "loss": 1.9635,
+      "step": 405
+    },
+    {
+      "epoch": 0.12691466083150985,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001995270142402049,
+      "loss": 1.788,
+      "step": 406
+    },
+    {
+      "epoch": 0.12722725851828695,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019952462430388506,
+      "loss": 1.7256,
+      "step": 407
+    },
+    {
+      "epoch": 0.12753985620506408,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019952222835913682,
+      "loss": 1.8476,
+      "step": 408
+    },
+    {
+      "epoch": 0.1278524538918412,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019951982640610484,
+      "loss": 1.9212,
+      "step": 409
+    },
+    {
+      "epoch": 0.12816505157861832,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019951741844493413,
+      "loss": 1.807,
+      "step": 410
+    },
+    {
+      "epoch": 0.12847764926539543,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019951500447577003,
+      "loss": 1.6015,
+      "step": 411
+    },
+    {
+      "epoch": 0.12879024695217256,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019951258449875828,
+      "loss": 1.8802,
+      "step": 412
+    },
+    {
+      "epoch": 0.12910284463894967,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019951015851404504,
+      "loss": 1.9614,
+      "step": 413
+    },
+    {
+      "epoch": 0.1294154423257268,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001995077265217767,
+      "loss": 1.8907,
+      "step": 414
+    },
+    {
+      "epoch": 0.1297280400125039,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019950528852210014,
+      "loss": 1.8123,
+      "step": 415
+    },
+    {
+      "epoch": 0.13004063769928104,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019950284451516245,
+      "loss": 1.6966,
+      "step": 416
+    },
+    {
+      "epoch": 0.13035323538605814,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019950039450111127,
+      "loss": 2.0439,
+      "step": 417
+    },
+    {
+      "epoch": 0.13066583307283527,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019949793848009448,
+      "loss": 1.9781,
+      "step": 418
+    },
+    {
+      "epoch": 0.13097843075961238,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019949547645226035,
+      "loss": 1.9264,
+      "step": 419
+    },
+    {
+      "epoch": 0.13129102844638948,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019949300841775753,
+      "loss": 2.0297,
+      "step": 420
+    },
+    {
+      "epoch": 0.13160362613316662,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.000199490534376735,
+      "loss": 1.9136,
+      "step": 421
+    },
+    {
+      "epoch": 0.13191622381994372,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019948805432934213,
+      "loss": 1.8224,
+      "step": 422
+    },
+    {
+      "epoch": 0.13222882150672086,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019948556827572862,
+      "loss": 1.7871,
+      "step": 423
+    },
+    {
+      "epoch": 0.13254141919349796,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019948307621604457,
+      "loss": 1.7048,
+      "step": 424
+    },
+    {
+      "epoch": 0.1328540168802751,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019948057815044048,
+      "loss": 1.9041,
+      "step": 425
+    },
+    {
+      "epoch": 0.1331666145670522,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001994780740790671,
+      "loss": 1.7443,
+      "step": 426
+    },
+    {
+      "epoch": 0.13347921225382933,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001994755640020756,
+      "loss": 1.6474,
+      "step": 427
+    },
+    {
+      "epoch": 0.13379180994060644,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019947304791961758,
+      "loss": 1.8303,
+      "step": 428
+    },
+    {
+      "epoch": 0.13410440762738357,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019947052583184488,
+      "loss": 1.64,
+      "step": 429
+    },
+    {
+      "epoch": 0.13441700531416068,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019946799773890974,
+      "loss": 1.7586,
+      "step": 430
+    },
+    {
+      "epoch": 0.13472960300093778,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019946546364096488,
+      "loss": 1.8402,
+      "step": 431
+    },
+    {
+      "epoch": 0.13504220068771491,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019946292353816318,
+      "loss": 2.2409,
+      "step": 432
+    },
+    {
+      "epoch": 0.13535479837449202,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001994603774306581,
+      "loss": 1.8416,
+      "step": 433
+    },
+    {
+      "epoch": 0.13566739606126915,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019945782531860325,
+      "loss": 1.7372,
+      "step": 434
+    },
+    {
+      "epoch": 0.13597999374804626,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019945526720215273,
+      "loss": 1.9704,
+      "step": 435
+    },
+    {
+      "epoch": 0.1362925914348234,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019945270308146103,
+      "loss": 1.6651,
+      "step": 436
+    },
+    {
+      "epoch": 0.1366051891216005,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019945013295668288,
+      "loss": 1.7958,
+      "step": 437
+    },
+    {
+      "epoch": 0.13691778680837763,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001994475568279735,
+      "loss": 2.0826,
+      "step": 438
+    },
+    {
+      "epoch": 0.13723038449515473,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019944497469548837,
+      "loss": 1.8808,
+      "step": 439
+    },
+    {
+      "epoch": 0.13754298218193187,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019944238655938339,
+      "loss": 2.257,
+      "step": 440
+    },
+    {
+      "epoch": 0.13785557986870897,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001994397924198148,
+      "loss": 2.0791,
+      "step": 441
+    },
+    {
+      "epoch": 0.13816817755548608,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019943719227693928,
+      "loss": 1.8917,
+      "step": 442
+    },
+    {
+      "epoch": 0.1384807752422632,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001994345861309137,
+      "loss": 1.8261,
+      "step": 443
+    },
+    {
+      "epoch": 0.13879337292904032,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019943197398189546,
+      "loss": 1.626,
+      "step": 444
+    },
+    {
+      "epoch": 0.13910597061581745,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019942935583004223,
+      "loss": 1.7819,
+      "step": 445
+    },
+    {
+      "epoch": 0.13941856830259455,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001994267316755121,
+      "loss": 1.8149,
+      "step": 446
+    },
+    {
+      "epoch": 0.1397311659893717,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019942410151846347,
+      "loss": 1.9703,
+      "step": 447
+    },
+    {
+      "epoch": 0.1400437636761488,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019942146535905514,
+      "loss": 1.7519,
+      "step": 448
+    },
+    {
+      "epoch": 0.14035636136292592,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019941882319744625,
+      "loss": 1.8088,
+      "step": 449
+    },
+    {
+      "epoch": 0.14066895904970303,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001994161750337963,
+      "loss": 2.0352,
+      "step": 450
+    },
+    {
+      "epoch": 0.14098155673648016,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001994135208682652,
+      "loss": 1.7832,
+      "step": 451
+    },
+    {
+      "epoch": 0.14129415442325727,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019941086070101314,
+      "loss": 1.7351,
+      "step": 452
+    },
+    {
+      "epoch": 0.14160675211003437,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019940819453220074,
+      "loss": 1.9127,
+      "step": 453
+    },
+    {
+      "epoch": 0.1419193497968115,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019940552236198897,
+      "loss": 2.6953,
+      "step": 454
+    },
+    {
+      "epoch": 0.1422319474835886,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019940284419053914,
+      "loss": 2.0053,
+      "step": 455
+    },
+    {
+      "epoch": 0.14254454517036574,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019940016001801294,
+      "loss": 1.7283,
+      "step": 456
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001993974698445724,
+      "loss": 1.7655,
+      "step": 457
+    },
+    {
+      "epoch": 0.14316974054391998,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019939477367037994,
+      "loss": 1.8373,
+      "step": 458
+    },
+    {
+      "epoch": 0.1434823382306971,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019939207149559835,
+      "loss": 1.8626,
+      "step": 459
+    },
+    {
+      "epoch": 0.14379493591747422,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019938936332039077,
+      "loss": 1.6125,
+      "step": 460
+    },
+    {
+      "epoch": 0.14410753360425133,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019938664914492062,
+      "loss": 2.0307,
+      "step": 461
+    },
+    {
+      "epoch": 0.14442013129102846,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019938392896935183,
+      "loss": 1.84,
+      "step": 462
+    },
+    {
+      "epoch": 0.14473272897780556,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001993812027938486,
+      "loss": 1.9634,
+      "step": 463
+    },
+    {
+      "epoch": 0.14504532666458267,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019937847061857552,
+      "loss": 2.0152,
+      "step": 464
+    },
+    {
+      "epoch": 0.1453579243513598,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019937573244369753,
+      "loss": 1.8692,
+      "step": 465
+    },
+    {
+      "epoch": 0.1456705220381369,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019937298826937995,
+      "loss": 1.7805,
+      "step": 466
+    },
+    {
+      "epoch": 0.14598311972491404,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019937023809578843,
+      "loss": 1.9569,
+      "step": 467
+    },
+    {
+      "epoch": 0.14629571741169115,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000199367481923089,
+      "loss": 1.9791,
+      "step": 468
+    },
+    {
+      "epoch": 0.14660831509846828,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019936471975144805,
+      "loss": 1.7193,
+      "step": 469
+    },
+    {
+      "epoch": 0.14692091278524538,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019936195158103237,
+      "loss": 1.7506,
+      "step": 470
+    },
+    {
+      "epoch": 0.14723351047202252,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019935917741200902,
+      "loss": 1.9867,
+      "step": 471
+    },
+    {
+      "epoch": 0.14754610815879962,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019935639724454556,
+      "loss": 1.8894,
+      "step": 472
+    },
+    {
+      "epoch": 0.14785870584557675,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019935361107880977,
+      "loss": 1.7917,
+      "step": 473
+    },
+    {
+      "epoch": 0.14817130353235386,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019935081891496985,
+      "loss": 1.9643,
+      "step": 474
+    },
+    {
+      "epoch": 0.14848390121913096,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001993480207531944,
+      "loss": 1.6624,
+      "step": 475
+    },
+    {
+      "epoch": 0.1487964989059081,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019934521659365235,
+      "loss": 1.5768,
+      "step": 476
+    },
+    {
+      "epoch": 0.1491090965926852,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019934240643651298,
+      "loss": 1.8556,
+      "step": 477
+    },
+    {
+      "epoch": 0.14942169427946234,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019933959028194592,
+      "loss": 1.9329,
+      "step": 478
+    },
+    {
+      "epoch": 0.14973429196623944,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001993367681301212,
+      "loss": 1.7054,
+      "step": 479
+    },
+    {
+      "epoch": 0.15004688965301657,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001993339399812092,
+      "loss": 1.8809,
+      "step": 480
+    },
+    {
+      "epoch": 0.15035948733979368,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001993311058353807,
+      "loss": 1.5983,
+      "step": 481
+    },
+    {
+      "epoch": 0.1506720850265708,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019932826569280673,
+      "loss": 1.7169,
+      "step": 482
+    },
+    {
+      "epoch": 0.15098468271334792,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019932541955365883,
+      "loss": 1.9345,
+      "step": 483
+    },
+    {
+      "epoch": 0.15129728040012505,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019932256741810874,
+      "loss": 2.1597,
+      "step": 484
+    },
+    {
+      "epoch": 0.15160987808690216,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001993197092863287,
+      "loss": 1.5661,
+      "step": 485
+    },
+    {
+      "epoch": 0.1519224757736793,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001993168451584912,
+      "loss": 1.8121,
+      "step": 486
+    },
+    {
+      "epoch": 0.1522350734604564,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019931397503476924,
+      "loss": 1.7365,
+      "step": 487
+    },
+    {
+      "epoch": 0.1525476711472335,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019931109891533605,
+      "loss": 1.6982,
+      "step": 488
+    },
+    {
+      "epoch": 0.15286026883401063,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019930821680036527,
+      "loss": 1.9638,
+      "step": 489
+    },
+    {
+      "epoch": 0.15317286652078774,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019930532869003086,
+      "loss": 2.1991,
+      "step": 490
+    },
+    {
+      "epoch": 0.15348546420756487,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019930243458450724,
+      "loss": 1.8095,
+      "step": 491
+    },
+    {
+      "epoch": 0.15379806189434198,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001992995344839691,
+      "loss": 1.9021,
+      "step": 492
+    },
+    {
+      "epoch": 0.1541106595811191,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001992966283885915,
+      "loss": 1.9448,
+      "step": 493
+    },
+    {
+      "epoch": 0.1544232572678962,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019929371629854992,
+      "loss": 1.9806,
+      "step": 494
+    },
+    {
+      "epoch": 0.15473585495467335,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001992907982140202,
+      "loss": 1.7495,
+      "step": 495
+    },
+    {
+      "epoch": 0.15504845264145045,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019928787413517842,
+      "loss": 2.0022,
+      "step": 496
+    },
+    {
+      "epoch": 0.15536105032822758,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019928494406220115,
+      "loss": 1.7185,
+      "step": 497
+    },
+    {
+      "epoch": 0.1556736480150047,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019928200799526532,
+      "loss": 2.0288,
+      "step": 498
+    },
+    {
+      "epoch": 0.1559862457017818,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019927906593454812,
+      "loss": 1.7969,
+      "step": 499
+    },
+    {
+      "epoch": 0.15629884338855893,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001992761178802272,
+      "loss": 2.1816,
+      "step": 500
+    },
+    {
+      "epoch": 0.15661144107533603,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019927316383248054,
+      "loss": 1.8524,
+      "step": 501
+    },
+    {
+      "epoch": 0.15692403876211317,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019927020379148646,
+      "loss": 1.6543,
+      "step": 502
+    },
+    {
+      "epoch": 0.15723663644889027,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001992672377574237,
+      "loss": 1.7662,
+      "step": 503
+    },
+    {
+      "epoch": 0.1575492341356674,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001992642657304713,
+      "loss": 1.8305,
+      "step": 504
+    },
+    {
+      "epoch": 0.1578618318224445,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019926128771080868,
+      "loss": 1.6887,
+      "step": 505
+    },
+    {
+      "epoch": 0.15817442950922164,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019925830369861564,
+      "loss": 1.9668,
+      "step": 506
+    },
+    {
+      "epoch": 0.15848702719599875,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019925531369407228,
+      "loss": 1.8739,
+      "step": 507
+    },
+    {
+      "epoch": 0.15879962488277588,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019925231769735917,
+      "loss": 1.8289,
+      "step": 508
+    },
+    {
+      "epoch": 0.15911222256955299,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001992493157086572,
+      "loss": 1.9057,
+      "step": 509
+    },
+    {
+      "epoch": 0.1594248202563301,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019924630772814753,
+      "loss": 1.8643,
+      "step": 510
+    },
+    {
+      "epoch": 0.15973741794310722,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019924329375601177,
+      "loss": 1.8911,
+      "step": 511
+    },
+    {
+      "epoch": 0.16005001562988433,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019924027379243192,
+      "loss": 1.6922,
+      "step": 512
+    },
+    {
+      "epoch": 0.16036261331666146,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001992372478375903,
+      "loss": 1.9621,
+      "step": 513
+    },
+    {
+      "epoch": 0.16067521100343857,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019923421589166954,
+      "loss": 1.8731,
+      "step": 514
+    },
+    {
+      "epoch": 0.1609878086902157,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019923117795485272,
+      "loss": 1.6659,
+      "step": 515
+    },
+    {
+      "epoch": 0.1613004063769928,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019922813402732325,
+      "loss": 1.9896,
+      "step": 516
+    },
+    {
+      "epoch": 0.16161300406376994,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019922508410926489,
+      "loss": 1.8087,
+      "step": 517
+    },
+    {
+      "epoch": 0.16192560175054704,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019922202820086171,
+      "loss": 2.0338,
+      "step": 518
+    },
+    {
+      "epoch": 0.16223819943732418,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019921896630229827,
+      "loss": 1.8984,
+      "step": 519
+    },
+    {
+      "epoch": 0.16255079712410128,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001992158984137594,
+      "loss": 1.7892,
+      "step": 520
+    },
+    {
+      "epoch": 0.1628633948108784,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019921282453543032,
+      "loss": 1.6763,
+      "step": 521
+    },
+    {
+      "epoch": 0.16317599249765552,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001992097446674966,
+      "loss": 1.8474,
+      "step": 522
+    },
+    {
+      "epoch": 0.16348859018443263,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019920665881014416,
+      "loss": 1.9876,
+      "step": 523
+    },
+    {
+      "epoch": 0.16380118787120976,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001992035669635593,
+      "loss": 1.7454,
+      "step": 524
+    },
+    {
+      "epoch": 0.16411378555798686,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001992004691279287,
+      "loss": 1.9164,
+      "step": 525
+    },
+    {
+      "epoch": 0.164426383244764,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019919736530343935,
+      "loss": 1.9096,
+      "step": 526
+    },
+    {
+      "epoch": 0.1647389809315411,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019919425549027865,
+      "loss": 1.9148,
+      "step": 527
+    },
+    {
+      "epoch": 0.16505157861831823,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019919113968863437,
+      "loss": 1.9967,
+      "step": 528
+    },
+    {
+      "epoch": 0.16536417630509534,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019918801789869453,
+      "loss": 1.9329,
+      "step": 529
+    },
+    {
+      "epoch": 0.16567677399187247,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019918489012064772,
+      "loss": 1.9399,
+      "step": 530
+    },
+    {
+      "epoch": 0.16598937167864958,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019918175635468265,
+      "loss": 1.9082,
+      "step": 531
+    },
+    {
+      "epoch": 0.16630196936542668,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019917861660098858,
+      "loss": 1.9138,
+      "step": 532
+    },
+    {
+      "epoch": 0.16661456705220382,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019917547085975505,
+      "loss": 1.7534,
+      "step": 533
+    },
+    {
+      "epoch": 0.16692716473898092,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019917231913117197,
+      "loss": 1.8574,
+      "step": 534
+    },
+    {
+      "epoch": 0.16723976242575805,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001991691614154296,
+      "loss": 1.7967,
+      "step": 535
+    },
+    {
+      "epoch": 0.16755236011253516,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019916599771271855,
+      "loss": 1.765,
+      "step": 536
+    },
+    {
+      "epoch": 0.1678649577993123,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019916282802322989,
+      "loss": 1.9999,
+      "step": 537
+    },
+    {
+      "epoch": 0.1681775554860894,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019915965234715491,
+      "loss": 1.9353,
+      "step": 538
+    },
+    {
+      "epoch": 0.16849015317286653,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019915647068468538,
+      "loss": 1.8003,
+      "step": 539
+    },
+    {
+      "epoch": 0.16880275085964364,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019915328303601334,
+      "loss": 2.1542,
+      "step": 540
+    },
+    {
+      "epoch": 0.16911534854642077,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019915008940133127,
+      "loss": 1.9446,
+      "step": 541
+    },
+    {
+      "epoch": 0.16942794623319787,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019914688978083192,
+      "loss": 2.0184,
+      "step": 542
+    },
+    {
+      "epoch": 0.16974054391997498,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019914368417470852,
+      "loss": 1.8707,
+      "step": 543
+    },
+    {
+      "epoch": 0.1700531416067521,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019914047258315457,
+      "loss": 1.8503,
+      "step": 544
+    },
+    {
+      "epoch": 0.17036573929352922,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019913725500636393,
+      "loss": 1.9382,
+      "step": 545
+    },
+    {
+      "epoch": 0.17067833698030635,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019913403144453088,
+      "loss": 1.6436,
+      "step": 546
+    },
+    {
+      "epoch": 0.17099093466708346,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019913080189785002,
+      "loss": 2.0155,
+      "step": 547
+    },
+    {
+      "epoch": 0.1713035323538606,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019912756636651638,
+      "loss": 1.9679,
+      "step": 548
+    },
+    {
+      "epoch": 0.1716161300406377,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019912432485072516,
+      "loss": 1.619,
+      "step": 549
+    },
+    {
+      "epoch": 0.17192872772741483,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001991210773506722,
+      "loss": 1.8251,
+      "step": 550
+    },
+    {
+      "epoch": 0.17224132541419193,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019911782386655341,
+      "loss": 1.9356,
+      "step": 551
+    },
+    {
+      "epoch": 0.17255392310096906,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019911456439856536,
+      "loss": 1.7967,
+      "step": 552
+    },
+    {
+      "epoch": 0.17286652078774617,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019911129894690475,
+      "loss": 1.7887,
+      "step": 553
+    },
+    {
+      "epoch": 0.17317911847452327,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019910802751176867,
+      "loss": 1.8225,
+      "step": 554
+    },
+    {
+      "epoch": 0.1734917161613004,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019910475009335472,
+      "loss": 1.7761,
+      "step": 555
+    },
+    {
+      "epoch": 0.1738043138480775,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001991014666918607,
+      "loss": 1.917,
+      "step": 556
+    },
+    {
+      "epoch": 0.17411691153485465,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019909817730748487,
+      "loss": 1.707,
+      "step": 557
+    },
+    {
+      "epoch": 0.17442950922163175,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019909488194042575,
+      "loss": 2.2473,
+      "step": 558
+    },
+    {
+      "epoch": 0.17474210690840888,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019909158059088235,
+      "loss": 1.5952,
+      "step": 559
+    },
+    {
+      "epoch": 0.175054704595186,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.000199088273259054,
+      "loss": 1.6575,
+      "step": 560
+    },
+    {
+      "epoch": 0.17536730228196312,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019908495994514026,
+      "loss": 1.9749,
+      "step": 561
+    },
+    {
+      "epoch": 0.17567989996874023,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019908164064934126,
+      "loss": 1.681,
+      "step": 562
+    },
+    {
+      "epoch": 0.17599249765551736,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019907831537185734,
+      "loss": 1.7532,
+      "step": 563
+    },
+    {
+      "epoch": 0.17630509534229447,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019907498411288925,
+      "loss": 2.0639,
+      "step": 564
+    },
+    {
+      "epoch": 0.1766176930290716,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019907164687263813,
+      "loss": 2.1285,
+      "step": 565
+    },
+    {
+      "epoch": 0.1769302907158487,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019906830365130546,
+      "loss": 1.7988,
+      "step": 566
+    },
+    {
+      "epoch": 0.1772428884026258,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019906495444909302,
+      "loss": 1.6593,
+      "step": 567
+    },
+    {
+      "epoch": 0.17755548608940294,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019906159926620306,
+      "loss": 1.8094,
+      "step": 568
+    },
+    {
+      "epoch": 0.17786808377618005,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019905823810283812,
+      "loss": 1.6249,
+      "step": 569
+    },
+    {
+      "epoch": 0.17818068146295718,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001990548709592011,
+      "loss": 1.6268,
+      "step": 570
+    },
+    {
+      "epoch": 0.17849327914973429,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019905149783549532,
+      "loss": 1.5067,
+      "step": 571
+    },
+    {
+      "epoch": 0.17880587683651142,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019904811873192437,
+      "loss": 1.7792,
+      "step": 572
+    },
+    {
+      "epoch": 0.17911847452328852,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001990447336486923,
+      "loss": 1.7893,
+      "step": 573
+    },
+    {
+      "epoch": 0.17943107221006566,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001990413425860034,
+      "loss": 1.7304,
+      "step": 574
+    },
+    {
+      "epoch": 0.17974366989684276,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019903794554406248,
+      "loss": 1.9092,
+      "step": 575
+    },
+    {
+      "epoch": 0.1800562675836199,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019903454252307454,
+      "loss": 1.6916,
+      "step": 576
+    },
+    {
+      "epoch": 0.180368865270397,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001990311335232451,
+      "loss": 1.7437,
+      "step": 577
+    },
+    {
+      "epoch": 0.1806814629571741,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019902771854477994,
+      "loss": 1.7296,
+      "step": 578
+    },
+    {
+      "epoch": 0.18099406064395124,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001990242975878852,
+      "loss": 1.6257,
+      "step": 579
+    },
+    {
+      "epoch": 0.18130665833072834,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001990208706527674,
+      "loss": 1.6635,
+      "step": 580
+    },
+    {
+      "epoch": 0.18161925601750548,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019901743773963353,
+      "loss": 1.8428,
+      "step": 581
+    },
+    {
+      "epoch": 0.18193185370428258,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019901399884869072,
+      "loss": 1.7945,
+      "step": 582
+    },
+    {
+      "epoch": 0.18224445139105971,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019901055398014662,
+      "loss": 1.7858,
+      "step": 583
+    },
+    {
+      "epoch": 0.18255704907783682,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001990071031342092,
+      "loss": 1.62,
+      "step": 584
+    },
+    {
+      "epoch": 0.18286964676461395,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019900364631108682,
+      "loss": 1.8136,
+      "step": 585
+    },
+    {
+      "epoch": 0.18318224445139106,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019900018351098813,
+      "loss": 1.9074,
+      "step": 586
+    },
+    {
+      "epoch": 0.1834948421381682,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001989967147341222,
+      "loss": 1.8761,
+      "step": 587
+    },
+    {
+      "epoch": 0.1838074398249453,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019899323998069846,
+      "loss": 1.8516,
+      "step": 588
+    },
+    {
+      "epoch": 0.1841200375117224,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001989897592509267,
+      "loss": 1.7505,
+      "step": 589
+    },
+    {
+      "epoch": 0.18443263519849953,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019898627254501697,
+      "loss": 1.9066,
+      "step": 590
+    },
+    {
+      "epoch": 0.18474523288527664,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001989827798631799,
+      "loss": 1.926,
+      "step": 591
+    },
+    {
+      "epoch": 0.18505783057205377,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019897928120562623,
+      "loss": 1.9225,
+      "step": 592
+    },
+    {
+      "epoch": 0.18537042825883088,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019897577657256724,
+      "loss": 2.0965,
+      "step": 593
+    },
+    {
+      "epoch": 0.185683025945608,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019897226596421447,
+      "loss": 1.7195,
+      "step": 594
+    },
+    {
+      "epoch": 0.18599562363238512,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019896874938077992,
+      "loss": 1.8197,
+      "step": 595
+    },
+    {
+      "epoch": 0.18630822131916225,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001989652268224758,
+      "loss": 2.2171,
+      "step": 596
+    },
+    {
+      "epoch": 0.18662081900593935,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019896169828951488,
+      "loss": 1.8195,
+      "step": 597
+    },
+    {
+      "epoch": 0.1869334166927165,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019895816378211008,
+      "loss": 1.6969,
+      "step": 598
+    },
+    {
+      "epoch": 0.1872460143794936,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019895462330047484,
+      "loss": 1.8099,
+      "step": 599
+    },
+    {
+      "epoch": 0.1875586120662707,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019895107684482293,
+      "loss": 1.7597,
+      "step": 600
+    },
+    {
+      "epoch": 0.18787120975304783,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019894752441536838,
+      "loss": 1.7928,
+      "step": 601
+    },
+    {
+      "epoch": 0.18818380743982493,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019894396601232567,
+      "loss": 1.7385,
+      "step": 602
+    },
+    {
+      "epoch": 0.18849640512660207,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001989404016359097,
+      "loss": 1.7216,
+      "step": 603
+    },
+    {
+      "epoch": 0.18880900281337917,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019893683128633557,
+      "loss": 1.749,
+      "step": 604
+    },
+    {
+      "epoch": 0.1891216005001563,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019893325496381884,
+      "loss": 1.8708,
+      "step": 605
+    },
+    {
+      "epoch": 0.1894341981869334,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019892967266857547,
+      "loss": 1.9852,
+      "step": 606
+    },
+    {
+      "epoch": 0.18974679587371054,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001989260844008217,
+      "loss": 1.7595,
+      "step": 607
+    },
+    {
+      "epoch": 0.19005939356048765,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019892249016077412,
+      "loss": 1.7231,
+      "step": 608
+    },
+    {
+      "epoch": 0.19037199124726478,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001989188899486498,
+      "loss": 1.7735,
+      "step": 609
+    },
+    {
+      "epoch": 0.1906845889340419,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019891528376466598,
+      "loss": 1.8502,
+      "step": 610
+    },
+    {
+      "epoch": 0.190997186620819,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019891167160904046,
+      "loss": 1.8522,
+      "step": 611
+    },
+    {
+      "epoch": 0.19130978430759613,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001989080534819913,
+      "loss": 2.0308,
+      "step": 612
+    },
+    {
+      "epoch": 0.19162238199437323,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019890442938373686,
+      "loss": 1.7471,
+      "step": 613
+    },
+    {
+      "epoch": 0.19193497968115036,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000198900799314496,
+      "loss": 1.5426,
+      "step": 614
+    },
+    {
+      "epoch": 0.19224757736792747,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001988971632744879,
+      "loss": 2.0733,
+      "step": 615
+    },
+    {
+      "epoch": 0.1925601750547046,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019889352126393198,
+      "loss": 1.8229,
+      "step": 616
+    },
+    {
+      "epoch": 0.1928727727414817,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019888987328304817,
+      "loss": 1.9119,
+      "step": 617
+    },
+    {
+      "epoch": 0.19318537042825884,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001988862193320567,
+      "loss": 1.6569,
+      "step": 618
+    },
+    {
+      "epoch": 0.19349796811503595,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019888255941117816,
+      "loss": 2.0652,
+      "step": 619
+    },
+    {
+      "epoch": 0.19381056580181308,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001988788935206335,
+      "loss": 1.6115,
+      "step": 620
+    },
+    {
+      "epoch": 0.19412316348859018,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019887522166064402,
+      "loss": 1.6017,
+      "step": 621
+    },
+    {
+      "epoch": 0.1944357611753673,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019887154383143143,
+      "loss": 1.9108,
+      "step": 622
+    },
+    {
+      "epoch": 0.19474835886214442,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019886786003321772,
+      "loss": 1.6372,
+      "step": 623
+    },
+    {
+      "epoch": 0.19506095654892153,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0001988641702662253,
+      "loss": 2.4569,
+      "step": 624
+    },
+    {
+      "epoch": 0.19537355423569866,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000198860474530677,
+      "loss": 1.6954,
+      "step": 625
+    },
+    {
+      "epoch": 0.19568615192247577,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019885677282679585,
+      "loss": 1.8825,
+      "step": 626
+    },
+    {
+      "epoch": 0.1959987496092529,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019885306515480533,
+      "loss": 1.7887,
+      "step": 627
+    },
+    {
+      "epoch": 0.19631134729603,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019884935151492933,
+      "loss": 1.8936,
+      "step": 628
+    },
+    {
+      "epoch": 0.19662394498280714,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019884563190739196,
+      "loss": 1.7583,
+      "step": 629
+    },
+    {
+      "epoch": 0.19693654266958424,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001988419063324179,
+      "loss": 1.898,
+      "step": 630
+    },
+    {
+      "epoch": 0.19724914035636137,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001988381747902319,
+      "loss": 2.0045,
+      "step": 631
+    },
+    {
+      "epoch": 0.19756173804313848,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019883443728105943,
+      "loss": 1.9453,
+      "step": 632
+    },
+    {
+      "epoch": 0.1978743357299156,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000198830693805126,
+      "loss": 1.8481,
+      "step": 633
+    },
+    {
+      "epoch": 0.19818693341669272,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019882694436265764,
+      "loss": 1.8409,
+      "step": 634
+    },
+    {
+      "epoch": 0.19849953110346982,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019882318895388072,
+      "loss": 1.8232,
+      "step": 635
+    },
+    {
+      "epoch": 0.19881212879024696,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019881942757902197,
+      "loss": 1.7768,
+      "step": 636
+    },
+    {
+      "epoch": 0.19912472647702406,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001988156602383084,
+      "loss": 1.7056,
+      "step": 637
+    },
+    {
+      "epoch": 0.1994373241638012,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019881188693196756,
+      "loss": 1.5243,
+      "step": 638
+    },
+    {
+      "epoch": 0.1997499218505783,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019880810766022714,
+      "loss": 2.0564,
+      "step": 639
+    },
+    {
+      "epoch": 0.20006251953735543,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019880432242331536,
+      "loss": 1.8789,
+      "step": 640
+    },
+    {
+      "epoch": 0.20037511722413254,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019880053122146073,
+      "loss": 1.8037,
+      "step": 641
+    },
+    {
+      "epoch": 0.20068771491090967,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019879673405489215,
+      "loss": 1.7692,
+      "step": 642
+    },
+    {
+      "epoch": 0.20100031259768678,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019879293092383882,
+      "loss": 1.7066,
+      "step": 643
+    },
+    {
+      "epoch": 0.2013129102844639,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019878912182853036,
+      "loss": 1.8715,
+      "step": 644
+    },
+    {
+      "epoch": 0.201625507971241,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001987853067691967,
+      "loss": 1.6647,
+      "step": 645
+    },
+    {
+      "epoch": 0.20193810565801812,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019878148574606824,
+      "loss": 1.6027,
+      "step": 646
+    },
+    {
+      "epoch": 0.20225070334479525,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019877765875937558,
+      "loss": 1.6788,
+      "step": 647
+    },
+    {
+      "epoch": 0.20256330103157236,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019877382580934977,
+      "loss": 1.7934,
+      "step": 648
+    },
+    {
+      "epoch": 0.2028758987183495,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019876998689622225,
+      "loss": 1.6556,
+      "step": 649
+    },
+    {
+      "epoch": 0.2031884964051266,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019876614202022475,
+      "loss": 1.7103,
+      "step": 650
+    },
+    {
+      "epoch": 0.20350109409190373,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001987622911815894,
+      "loss": 1.7654,
+      "step": 651
+    },
+    {
+      "epoch": 0.20381369177868083,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019875843438054864,
+      "loss": 1.7043,
+      "step": 652
+    },
+    {
+      "epoch": 0.20412628946545797,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001987545716173354,
+      "loss": 1.966,
+      "step": 653
+    },
+    {
+      "epoch": 0.20443888715223507,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001987507028921828,
+      "loss": 1.7629,
+      "step": 654
+    },
+    {
+      "epoch": 0.2047514848390122,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019874682820532444,
+      "loss": 1.766,
+      "step": 655
+    },
+    {
+      "epoch": 0.2050640825257893,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019874294755699423,
+      "loss": 1.6821,
+      "step": 656
+    },
+    {
+      "epoch": 0.20537668021256641,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019873906094742644,
+      "loss": 1.806,
+      "step": 657
+    },
+    {
+      "epoch": 0.20568927789934355,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001987351683768557,
+      "loss": 1.8864,
+      "step": 658
+    },
+    {
+      "epoch": 0.20600187558612065,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019873126984551703,
+      "loss": 1.7406,
+      "step": 659
+    },
+    {
+      "epoch": 0.20631447327289779,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001987273653536458,
+      "loss": 1.7246,
+      "step": 660
+    },
+    {
+      "epoch": 0.2066270709596749,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019872345490147772,
+      "loss": 1.9874,
+      "step": 661
+    },
+    {
+      "epoch": 0.20693966864645202,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019871953848924886,
+      "loss": 1.7792,
+      "step": 662
+    },
+    {
+      "epoch": 0.20725226633322913,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019871561611719564,
+      "loss": 1.8759,
+      "step": 663
+    },
+    {
+      "epoch": 0.20756486402000626,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019871168778555492,
+      "loss": 1.9906,
+      "step": 664
+    },
+    {
+      "epoch": 0.20787746170678337,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001987077534945638,
+      "loss": 1.8973,
+      "step": 665
+    },
+    {
+      "epoch": 0.2081900593935605,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019870381324445978,
+      "loss": 1.6312,
+      "step": 666
+    },
+    {
+      "epoch": 0.2085026570803376,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001986998670354808,
+      "loss": 1.8406,
+      "step": 667
+    },
+    {
+      "epoch": 0.2088152547671147,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001986959148678651,
+      "loss": 1.7828,
+      "step": 668
+    },
+    {
+      "epoch": 0.20912785245389184,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019869195674185122,
+      "loss": 1.9185,
+      "step": 669
+    },
+    {
+      "epoch": 0.20944045014066895,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019868799265767816,
+      "loss": 1.7588,
+      "step": 670
+    },
+    {
+      "epoch": 0.20975304782744608,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019868402261558524,
+      "loss": 1.7387,
+      "step": 671
+    },
+    {
+      "epoch": 0.2100656455142232,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019868004661581208,
+      "loss": 1.6164,
+      "step": 672
+    },
+    {
+      "epoch": 0.21037824320100032,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001986760646585988,
+      "loss": 1.8667,
+      "step": 673
+    },
+    {
+      "epoch": 0.21069084088777743,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019867207674418568,
+      "loss": 1.9312,
+      "step": 674
+    },
+    {
+      "epoch": 0.21100343857455456,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001986680828728136,
+      "loss": 1.7665,
+      "step": 675
+    },
+    {
+      "epoch": 0.21131603626133166,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019866408304472364,
+      "loss": 1.6056,
+      "step": 676
+    },
+    {
+      "epoch": 0.2116286339481088,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019866007726015723,
+      "loss": 1.5752,
+      "step": 677
+    },
+    {
+      "epoch": 0.2119412316348859,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019865606551935626,
+      "loss": 1.8815,
+      "step": 678
+    },
+    {
+      "epoch": 0.212253829321663,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019865204782256287,
+      "loss": 1.7828,
+      "step": 679
+    },
+    {
+      "epoch": 0.21256642700844014,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001986480241700196,
+      "loss": 1.9457,
+      "step": 680
+    },
+    {
+      "epoch": 0.21287902469521724,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019864399456196946,
+      "loss": 1.9523,
+      "step": 681
+    },
+    {
+      "epoch": 0.21319162238199438,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019863995899865565,
+      "loss": 1.5974,
+      "step": 682
+    },
+    {
+      "epoch": 0.21350422006877148,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019863591748032184,
+      "loss": 1.8886,
+      "step": 683
+    },
+    {
+      "epoch": 0.21381681775554862,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019863187000721197,
+      "loss": 1.8564,
+      "step": 684
+    },
+    {
+      "epoch": 0.21412941544232572,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019862781657957045,
+      "loss": 1.8022,
+      "step": 685
+    },
+    {
+      "epoch": 0.21444201312910285,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019862375719764192,
+      "loss": 1.855,
+      "step": 686
+    },
+    {
+      "epoch": 0.21475461081587996,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001986196918616715,
+      "loss": 2.0019,
+      "step": 687
+    },
+    {
+      "epoch": 0.2150672085026571,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019861562057190462,
+      "loss": 1.8597,
+      "step": 688
+    },
+    {
+      "epoch": 0.2153798061894342,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019861154332858708,
+      "loss": 1.9685,
+      "step": 689
+    },
+    {
+      "epoch": 0.2156924038762113,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019860746013196495,
+      "loss": 1.8702,
+      "step": 690
+    },
+    {
+      "epoch": 0.21600500156298844,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019860337098228485,
+      "loss": 1.6556,
+      "step": 691
+    },
+    {
+      "epoch": 0.21631759924976554,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019859927587979358,
+      "loss": 2.0366,
+      "step": 692
+    },
+    {
+      "epoch": 0.21663019693654267,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019859517482473838,
+      "loss": 1.9303,
+      "step": 693
+    },
+    {
+      "epoch": 0.21694279462331978,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019859106781736682,
+      "loss": 1.6981,
+      "step": 694
+    },
+    {
+      "epoch": 0.2172553923100969,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019858695485792686,
+      "loss": 1.4825,
+      "step": 695
+    },
+    {
+      "epoch": 0.21756798999687402,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001985828359466668,
+      "loss": 1.779,
+      "step": 696
+    },
+    {
+      "epoch": 0.21788058768365115,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019857871108383532,
+      "loss": 1.7535,
+      "step": 697
+    },
+    {
+      "epoch": 0.21819318537042826,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019857458026968143,
+      "loss": 1.7039,
+      "step": 698
+    },
+    {
+      "epoch": 0.2185057830572054,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001985704435044545,
+      "loss": 1.7501,
+      "step": 699
+    },
+    {
+      "epoch": 0.2188183807439825,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001985663007884043,
+      "loss": 1.8791,
+      "step": 700
+    },
+    {
+      "epoch": 0.2191309784307596,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019856215212178094,
+      "loss": 1.904,
+      "step": 701
+    },
+    {
+      "epoch": 0.21944357611753673,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019855799750483484,
+      "loss": 1.4772,
+      "step": 702
+    },
+    {
+      "epoch": 0.21975617380431384,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019855383693781682,
+      "loss": 2.4316,
+      "step": 703
+    },
+    {
+      "epoch": 0.22006877149109097,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001985496704209781,
+      "loss": 1.6331,
+      "step": 704
+    },
+    {
+      "epoch": 0.22038136917786808,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001985454979545702,
+      "loss": 1.7665,
+      "step": 705
+    },
+    {
+      "epoch": 0.2206939668646452,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019854131953884495,
+      "loss": 1.9052,
+      "step": 706
+    },
+    {
+      "epoch": 0.2210065645514223,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019853713517405472,
+      "loss": 1.8316,
+      "step": 707
+    },
+    {
+      "epoch": 0.22131916223819945,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019853294486045208,
+      "loss": 1.6123,
+      "step": 708
+    },
+    {
+      "epoch": 0.22163175992497655,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019852874859828997,
+      "loss": 1.8111,
+      "step": 709
+    },
+    {
+      "epoch": 0.22194435761175368,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019852454638782176,
+      "loss": 1.8234,
+      "step": 710
+    },
+    {
+      "epoch": 0.2222569552985308,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019852033822930114,
+      "loss": 1.6664,
+      "step": 711
+    },
+    {
+      "epoch": 0.22256955298530792,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019851612412298214,
+      "loss": 1.9896,
+      "step": 712
+    },
+    {
+      "epoch": 0.22288215067208503,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001985119040691192,
+      "loss": 1.6152,
+      "step": 713
+    },
+    {
+      "epoch": 0.22319474835886213,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019850767806796707,
+      "loss": 2.165,
+      "step": 714
+    },
+    {
+      "epoch": 0.22350734604563927,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019850344611978087,
+      "loss": 2.1852,
+      "step": 715
+    },
+    {
+      "epoch": 0.22381994373241637,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019849920822481614,
+      "loss": 1.7914,
+      "step": 716
+    },
+    {
+      "epoch": 0.2241325414191935,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019849496438332866,
+      "loss": 2.0296,
+      "step": 717
+    },
+    {
+      "epoch": 0.2244451391059706,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001984907145955747,
+      "loss": 1.7981,
+      "step": 718
+    },
+    {
+      "epoch": 0.22475773679274774,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019848645886181074,
+      "loss": 1.7928,
+      "step": 719
+    },
+    {
+      "epoch": 0.22507033447952485,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019848219718229378,
+      "loss": 1.8671,
+      "step": 720
+    },
+    {
+      "epoch": 0.22538293216630198,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019847792955728107,
+      "loss": 1.8564,
+      "step": 721
+    },
+    {
+      "epoch": 0.22569552985307909,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001984736559870303,
+      "loss": 1.6293,
+      "step": 722
+    },
+    {
+      "epoch": 0.22600812753985622,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001984693764717994,
+      "loss": 1.9545,
+      "step": 723
+    },
+    {
+      "epoch": 0.22632072522663332,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019846509101184679,
+      "loss": 1.8173,
+      "step": 724
+    },
+    {
+      "epoch": 0.22663332291341043,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019846079960743112,
+      "loss": 1.649,
+      "step": 725
+    },
+    {
+      "epoch": 0.22694592060018756,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019845650225881154,
+      "loss": 1.8916,
+      "step": 726
+    },
+    {
+      "epoch": 0.22725851828696467,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019845219896624743,
+      "loss": 1.7553,
+      "step": 727
+    },
+    {
+      "epoch": 0.2275711159737418,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001984478897299986,
+      "loss": 1.6849,
+      "step": 728
+    },
+    {
+      "epoch": 0.2278837136605189,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019844357455032526,
+      "loss": 1.8667,
+      "step": 729
+    },
+    {
+      "epoch": 0.22819631134729604,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019843925342748783,
+      "loss": 1.651,
+      "step": 730
+    },
+    {
+      "epoch": 0.22850890903407314,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019843492636174728,
+      "loss": 1.6074,
+      "step": 731
+    },
+    {
+      "epoch": 0.22882150672085028,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00019843059335336474,
+      "loss": 1.8431,
+      "step": 732
+    },
+    {
+      "epoch": 0.22913410440762738,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019842625440260188,
+      "loss": 1.7872,
+      "step": 733
+    },
+    {
+      "epoch": 0.22944670209440451,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001984219095097206,
+      "loss": 1.6808,
+      "step": 734
+    },
+    {
+      "epoch": 0.22975929978118162,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019841755867498322,
+      "loss": 1.7461,
+      "step": 735
+    },
+    {
+      "epoch": 0.23007189746795872,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019841320189865243,
+      "loss": 1.7838,
+      "step": 736
+    },
+    {
+      "epoch": 0.23038449515473586,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001984088391809912,
+      "loss": 1.9104,
+      "step": 737
+    },
+    {
+      "epoch": 0.23069709284151296,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019840447052226298,
+      "loss": 1.9961,
+      "step": 738
+    },
+    {
+      "epoch": 0.2310096905282901,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019840009592273143,
+      "loss": 1.987,
+      "step": 739
+    },
+    {
+      "epoch": 0.2313222882150672,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019839571538266072,
+      "loss": 1.6382,
+      "step": 740
+    },
+    {
+      "epoch": 0.23163488590184433,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001983913289023153,
+      "loss": 1.6738,
+      "step": 741
+    },
+    {
+      "epoch": 0.23194748358862144,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019838693648195995,
+      "loss": 1.8182,
+      "step": 742
+    },
+    {
+      "epoch": 0.23226008127539857,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019838253812185988,
+      "loss": 1.598,
+      "step": 743
+    },
+    {
+      "epoch": 0.23257267896217568,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019837813382228063,
+      "loss": 1.7465,
+      "step": 744
+    },
+    {
+      "epoch": 0.2328852766489528,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019837372358348806,
+      "loss": 1.8831,
+      "step": 745
+    },
+    {
+      "epoch": 0.23319787433572992,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019836930740574845,
+      "loss": 1.525,
+      "step": 746
+    },
+    {
+      "epoch": 0.23351047202250702,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00019836488528932836,
+      "loss": 3.4084,
+      "step": 747
+    },
+    {
+      "epoch": 0.23382306970928415,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019836045723449483,
+      "loss": 1.7993,
+      "step": 748
+    },
+    {
+      "epoch": 0.23413566739606126,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019835602324151514,
+      "loss": 1.8971,
+      "step": 749
+    },
+    {
+      "epoch": 0.2344482650828384,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019835158331065703,
+      "loss": 1.7286,
+      "step": 750
+    },
+    {
+      "epoch": 0.2347608627696155,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019834713744218844,
+      "loss": 1.6018,
+      "step": 751
+    },
+    {
+      "epoch": 0.23507346045639263,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019834268563637787,
+      "loss": 1.8705,
+      "step": 752
+    },
+    {
+      "epoch": 0.23538605814316974,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019833822789349409,
+      "loss": 1.8121,
+      "step": 753
+    },
+    {
+      "epoch": 0.23569865582994687,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019833376421380612,
+      "loss": 1.6886,
+      "step": 754
+    },
+    {
+      "epoch": 0.23601125351672397,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019832929459758352,
+      "loss": 1.6922,
+      "step": 755
+    },
+    {
+      "epoch": 0.2363238512035011,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001983248190450961,
+      "loss": 1.7953,
+      "step": 756
+    },
+    {
+      "epoch": 0.2366364488902782,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019832033755661405,
+      "loss": 1.7892,
+      "step": 757
+    },
+    {
+      "epoch": 0.23694904657705532,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019831585013240793,
+      "loss": 1.5738,
+      "step": 758
+    },
+    {
+      "epoch": 0.23726164426383245,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001983113567727487,
+      "loss": 1.6719,
+      "step": 759
+    },
+    {
+      "epoch": 0.23757424195060955,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019830685747790748,
+      "loss": 1.9564,
+      "step": 760
+    },
+    {
+      "epoch": 0.2378868396373867,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001983023522481561,
+      "loss": 2.1432,
+      "step": 761
+    },
+    {
+      "epoch": 0.2381994373241638,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001982978410837664,
+      "loss": 1.8179,
+      "step": 762
+    },
+    {
+      "epoch": 0.23851203501094093,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001982933239850108,
+      "loss": 1.9499,
+      "step": 763
+    },
+    {
+      "epoch": 0.23882463269771803,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019828880095216193,
+      "loss": 1.5989,
+      "step": 764
+    },
+    {
+      "epoch": 0.23913723038449516,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019828427198549293,
+      "loss": 1.8161,
+      "step": 765
+    },
+    {
+      "epoch": 0.23944982807127227,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001982797370852772,
+      "loss": 1.9312,
+      "step": 766
+    },
+    {
+      "epoch": 0.2397624257580494,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019827519625178846,
+      "loss": 1.7275,
+      "step": 767
+    },
+    {
+      "epoch": 0.2400750234448265,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001982706494853009,
+      "loss": 2.0002,
+      "step": 768
+    },
+    {
+      "epoch": 0.2403876211316036,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019826609678608902,
+      "loss": 1.8021,
+      "step": 769
+    },
+    {
+      "epoch": 0.24070021881838075,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019826153815442763,
+      "loss": 1.5546,
+      "step": 770
+    },
+    {
+      "epoch": 0.24101281650515785,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.000198256973590592,
+      "loss": 1.5848,
+      "step": 771
+    },
+    {
+      "epoch": 0.24132541419193498,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019825240309485765,
+      "loss": 1.7763,
+      "step": 772
+    },
+    {
+      "epoch": 0.2416380118787121,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001982478266675005,
+      "loss": 1.6906,
+      "step": 773
+    },
+    {
+      "epoch": 0.24195060956548922,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019824324430879687,
+      "loss": 1.9644,
+      "step": 774
+    },
+    {
+      "epoch": 0.24226320725226633,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019823865601902341,
+      "loss": 1.9122,
+      "step": 775
+    },
+    {
+      "epoch": 0.24257580493904346,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019823406179845707,
+      "loss": 1.9017,
+      "step": 776
+    },
+    {
+      "epoch": 0.24288840262582057,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019822946164737526,
+      "loss": 1.8361,
+      "step": 777
+    },
+    {
+      "epoch": 0.2432010003125977,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019822485556605566,
+      "loss": 1.7349,
+      "step": 778
+    },
+    {
+      "epoch": 0.2435135979993748,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019822024355477637,
+      "loss": 1.6017,
+      "step": 779
+    },
+    {
+      "epoch": 0.2438261956861519,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001982156256138158,
+      "loss": 1.8296,
+      "step": 780
+    },
+    {
+      "epoch": 0.24413879337292904,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019821100174345277,
+      "loss": 1.6754,
+      "step": 781
+    },
+    {
+      "epoch": 0.24445139105970615,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001982063719439664,
+      "loss": 2.0037,
+      "step": 782
+    },
+    {
+      "epoch": 0.24476398874648328,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019820173621563623,
+      "loss": 1.887,
+      "step": 783
+    },
+    {
+      "epoch": 0.24507658643326038,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001981970945587421,
+      "loss": 1.5708,
+      "step": 784
+    },
+    {
+      "epoch": 0.24538918412003752,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001981924469735642,
+      "loss": 2.3282,
+      "step": 785
+    },
+    {
+      "epoch": 0.24570178180681462,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019818779346038318,
+      "loss": 1.7515,
+      "step": 786
+    },
+    {
+      "epoch": 0.24601437949359176,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019818313401947997,
+      "loss": 1.7623,
+      "step": 787
+    },
+    {
+      "epoch": 0.24632697718036886,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019817846865113577,
+      "loss": 1.8036,
+      "step": 788
+    },
+    {
+      "epoch": 0.246639574867146,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001981737973556324,
+      "loss": 1.8455,
+      "step": 789
+    },
+    {
+      "epoch": 0.2469521725539231,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001981691201332517,
+      "loss": 1.7791,
+      "step": 790
+    },
+    {
+      "epoch": 0.24726477024070023,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019816443698427615,
+      "loss": 2.0416,
+      "step": 791
+    },
+    {
+      "epoch": 0.24757736792747734,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019815974790898846,
+      "loss": 2.2271,
+      "step": 792
+    },
+    {
+      "epoch": 0.24788996561425444,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019815505290767172,
+      "loss": 1.5433,
+      "step": 793
+    },
+    {
+      "epoch": 0.24820256330103158,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001981503519806093,
+      "loss": 1.7228,
+      "step": 794
+    },
+    {
+      "epoch": 0.24851516098780868,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019814564512808512,
+      "loss": 1.8217,
+      "step": 795
+    },
+    {
+      "epoch": 0.2488277586745858,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019814093235038323,
+      "loss": 1.8205,
+      "step": 796
+    },
+    {
+      "epoch": 0.24914035636136292,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019813621364778817,
+      "loss": 1.8541,
+      "step": 797
+    },
+    {
+      "epoch": 0.24945295404814005,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001981314890205849,
+      "loss": 1.8656,
+      "step": 798
+    },
+    {
+      "epoch": 0.24976555173491716,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019812675846905855,
+      "loss": 1.809,
+      "step": 799
+    },
+    {
+      "epoch": 0.25007814942169426,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019812202199349476,
+      "loss": 2.0585,
+      "step": 800
+    },
+    {
+      "epoch": 0.2503907471084714,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019811727959417945,
+      "loss": 1.9492,
+      "step": 801
+    },
+    {
+      "epoch": 0.25070334479524853,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019811253127139896,
+      "loss": 1.8192,
+      "step": 802
+    },
+    {
+      "epoch": 0.25101594248202563,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001981077770254399,
+      "loss": 1.4981,
+      "step": 803
+    },
+    {
+      "epoch": 0.25132854016880274,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019810301685658935,
+      "loss": 1.8598,
+      "step": 804
+    },
+    {
+      "epoch": 0.25164113785557984,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019809825076513464,
+      "loss": 1.7946,
+      "step": 805
+    },
+    {
+      "epoch": 0.251953735542357,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019809347875136352,
+      "loss": 1.784,
+      "step": 806
+    },
+    {
+      "epoch": 0.2522663332291341,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019808870081556413,
+      "loss": 1.9401,
+      "step": 807
+    },
+    {
+      "epoch": 0.2525789309159112,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019808391695802483,
+      "loss": 2.0217,
+      "step": 808
+    },
+    {
+      "epoch": 0.2528915286026883,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001980791271790345,
+      "loss": 1.7354,
+      "step": 809
+    },
+    {
+      "epoch": 0.2532041262894655,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019807433147888225,
+      "loss": 2.1094,
+      "step": 810
+    },
+    {
+      "epoch": 0.2535167239762426,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019806952985785764,
+      "loss": 2.8019,
+      "step": 811
+    },
+    {
+      "epoch": 0.2538293216630197,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019806472231625056,
+      "loss": 1.554,
+      "step": 812
+    },
+    {
+      "epoch": 0.2541419193497968,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001980599088543512,
+      "loss": 1.7158,
+      "step": 813
+    },
+    {
+      "epoch": 0.2544545170365739,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019805508947245021,
+      "loss": 1.934,
+      "step": 814
+    },
+    {
+      "epoch": 0.25476711472335106,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001980502641708385,
+      "loss": 2.0267,
+      "step": 815
+    },
+    {
+      "epoch": 0.25507971241012817,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001980454329498074,
+      "loss": 1.6819,
+      "step": 816
+    },
+    {
+      "epoch": 0.2553923100969053,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019804059580964855,
+      "loss": 1.7279,
+      "step": 817
+    },
+    {
+      "epoch": 0.2557049077836824,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019803575275065404,
+      "loss": 1.6234,
+      "step": 818
+    },
+    {
+      "epoch": 0.25601750547045954,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001980309037731162,
+      "loss": 1.4631,
+      "step": 819
+    },
+    {
+      "epoch": 0.25633010315723664,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019802604887732774,
+      "loss": 1.7769,
+      "step": 820
+    },
+    {
+      "epoch": 0.25664270084401375,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019802118806358182,
+      "loss": 1.7928,
+      "step": 821
+    },
+    {
+      "epoch": 0.25695529853079085,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019801632133217189,
+      "loss": 1.639,
+      "step": 822
+    },
+    {
+      "epoch": 0.257267896217568,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001980114486833917,
+      "loss": 1.6918,
+      "step": 823
+    },
+    {
+      "epoch": 0.2575804939043451,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019800657011753548,
+      "loss": 1.8273,
+      "step": 824
+    },
+    {
+      "epoch": 0.2578930915911222,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001980016856348977,
+      "loss": 1.9625,
+      "step": 825
+    },
+    {
+      "epoch": 0.25820568927789933,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019799679523577332,
+      "loss": 1.741,
+      "step": 826
+    },
+    {
+      "epoch": 0.25851828696467644,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019799189892045748,
+      "loss": 2.0397,
+      "step": 827
+    },
+    {
+      "epoch": 0.2588308846514536,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019798699668924585,
+      "loss": 1.7246,
+      "step": 828
+    },
+    {
+      "epoch": 0.2591434823382307,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019798208854243437,
+      "loss": 1.5622,
+      "step": 829
+    },
+    {
+      "epoch": 0.2594560800250078,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019797717448031936,
+      "loss": 1.4121,
+      "step": 830
+    },
+    {
+      "epoch": 0.2597686777117849,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019797225450319744,
+      "loss": 1.6693,
+      "step": 831
+    },
+    {
+      "epoch": 0.2600812753985621,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001979673286113657,
+      "loss": 1.6021,
+      "step": 832
+    },
+    {
+      "epoch": 0.2603938730853392,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001979623968051215,
+      "loss": 1.9199,
+      "step": 833
+    },
+    {
+      "epoch": 0.2607064707721163,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019795745908476254,
+      "loss": 2.0403,
+      "step": 834
+    },
+    {
+      "epoch": 0.2610190684588934,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019795251545058694,
+      "loss": 1.8294,
+      "step": 835
+    },
+    {
+      "epoch": 0.26133166614567055,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019794756590289317,
+      "loss": 1.8625,
+      "step": 836
+    },
+    {
+      "epoch": 0.26164426383244765,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019794261044198003,
+      "loss": 1.8086,
+      "step": 837
+    },
+    {
+      "epoch": 0.26195686151922476,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001979376490681467,
+      "loss": 1.6601,
+      "step": 838
+    },
+    {
+      "epoch": 0.26226945920600186,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019793268178169268,
+      "loss": 1.5396,
+      "step": 839
+    },
+    {
+      "epoch": 0.26258205689277897,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019792770858291788,
+      "loss": 1.7095,
+      "step": 840
+    },
+    {
+      "epoch": 0.26289465457955613,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019792272947212253,
+      "loss": 1.8782,
+      "step": 841
+    },
+    {
+      "epoch": 0.26320725226633324,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019791774444960717,
+      "loss": 1.9358,
+      "step": 842
+    },
+    {
+      "epoch": 0.26351984995311034,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019791275351567286,
+      "loss": 1.7342,
+      "step": 843
+    },
+    {
+      "epoch": 0.26383244763988745,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019790775667062081,
+      "loss": 1.7846,
+      "step": 844
+    },
+    {
+      "epoch": 0.2641450453266646,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001979027539147527,
+      "loss": 1.845,
+      "step": 845
+    },
+    {
+      "epoch": 0.2644576430134417,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001978977452483706,
+      "loss": 1.9035,
+      "step": 846
+    },
+    {
+      "epoch": 0.2647702407002188,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001978927306717769,
+      "loss": 1.5762,
+      "step": 847
+    },
+    {
+      "epoch": 0.2650828383869959,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001978877101852743,
+      "loss": 2.1721,
+      "step": 848
+    },
+    {
+      "epoch": 0.26539543607377303,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019788268378916586,
+      "loss": 1.7108,
+      "step": 849
+    },
+    {
+      "epoch": 0.2657080337605502,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019787765148375508,
+      "loss": 2.5699,
+      "step": 850
+    },
+    {
+      "epoch": 0.2660206314473273,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019787261326934577,
+      "loss": 1.6568,
+      "step": 851
+    },
+    {
+      "epoch": 0.2663332291341044,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019786756914624208,
+      "loss": 1.8594,
+      "step": 852
+    },
+    {
+      "epoch": 0.2666458268208815,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019786251911474849,
+      "loss": 1.8597,
+      "step": 853
+    },
+    {
+      "epoch": 0.26695842450765866,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019785746317516994,
+      "loss": 2.0457,
+      "step": 854
+    },
+    {
+      "epoch": 0.26727102219443577,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019785240132781163,
+      "loss": 1.832,
+      "step": 855
+    },
+    {
+      "epoch": 0.2675836198812129,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019784733357297915,
+      "loss": 1.68,
+      "step": 856
+    },
+    {
+      "epoch": 0.26789621756799,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019784225991097848,
+      "loss": 1.8997,
+      "step": 857
+    },
+    {
+      "epoch": 0.26820881525476714,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019783718034211586,
+      "loss": 1.7594,
+      "step": 858
+    },
+    {
+      "epoch": 0.26852141294154425,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000197832094866698,
+      "loss": 1.7918,
+      "step": 859
+    },
+    {
+      "epoch": 0.26883401062832135,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019782700348503193,
+      "loss": 1.6616,
+      "step": 860
+    },
+    {
+      "epoch": 0.26914660831509846,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019782190619742495,
+      "loss": 1.8357,
+      "step": 861
+    },
+    {
+      "epoch": 0.26945920600187556,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001978168030041849,
+      "loss": 1.74,
+      "step": 862
+    },
+    {
+      "epoch": 0.2697718036886527,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019781169390561975,
+      "loss": 1.4934,
+      "step": 863
+    },
+    {
+      "epoch": 0.27008440137542983,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000197806578902038,
+      "loss": 1.6285,
+      "step": 864
+    },
+    {
+      "epoch": 0.27039699906220693,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019780145799374848,
+      "loss": 1.5881,
+      "step": 865
+    },
+    {
+      "epoch": 0.27070959674898404,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019779633118106028,
+      "loss": 1.714,
+      "step": 866
+    },
+    {
+      "epoch": 0.2710221944357612,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000197791198464283,
+      "loss": 1.9303,
+      "step": 867
+    },
+    {
+      "epoch": 0.2713347921225383,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001977860598437264,
+      "loss": 1.6095,
+      "step": 868
+    },
+    {
+      "epoch": 0.2716473898093154,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019778091531970072,
+      "loss": 1.7565,
+      "step": 869
+    },
+    {
+      "epoch": 0.2719599874960925,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019777576489251664,
+      "loss": 1.5668,
+      "step": 870
+    },
+    {
+      "epoch": 0.2722725851828696,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019777060856248504,
+      "loss": 1.6762,
+      "step": 871
+    },
+    {
+      "epoch": 0.2725851828696468,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019776544632991717,
+      "loss": 1.7808,
+      "step": 872
+    },
+    {
+      "epoch": 0.2728977805564239,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019776027819512474,
+      "loss": 1.8983,
+      "step": 873
+    },
+    {
+      "epoch": 0.273210378243201,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019775510415841977,
+      "loss": 1.837,
+      "step": 874
+    },
+    {
+      "epoch": 0.2735229759299781,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019774992422011452,
+      "loss": 1.7363,
+      "step": 875
+    },
+    {
+      "epoch": 0.27383557361675526,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019774473838052184,
+      "loss": 1.8509,
+      "step": 876
+    },
+    {
+      "epoch": 0.27414817130353236,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019773954663995476,
+      "loss": 1.8239,
+      "step": 877
+    },
+    {
+      "epoch": 0.27446076899030947,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019773434899872665,
+      "loss": 2.0633,
+      "step": 878
+    },
+    {
+      "epoch": 0.2747733666770866,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019772914545715135,
+      "loss": 2.0269,
+      "step": 879
+    },
+    {
+      "epoch": 0.27508596436386373,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019772393601554303,
+      "loss": 1.7389,
+      "step": 880
+    },
+    {
+      "epoch": 0.27539856205064084,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019771872067421615,
+      "loss": 2.0936,
+      "step": 881
+    },
+    {
+      "epoch": 0.27571115973741794,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019771349943348558,
+      "loss": 1.7132,
+      "step": 882
+    },
+    {
+      "epoch": 0.27602375742419505,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019770827229366654,
+      "loss": 1.6179,
+      "step": 883
+    },
+    {
+      "epoch": 0.27633635511097215,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019770303925507456,
+      "loss": 1.9907,
+      "step": 884
+    },
+    {
+      "epoch": 0.2766489527977493,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001976978003180256,
+      "loss": 1.5918,
+      "step": 885
+    },
+    {
+      "epoch": 0.2769615504845264,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000197692555482836,
+      "loss": 1.6069,
+      "step": 886
+    },
+    {
+      "epoch": 0.2772741481713035,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019768730474982227,
+      "loss": 1.9966,
+      "step": 887
+    },
+    {
+      "epoch": 0.27758674585808063,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019768204811930148,
+      "loss": 1.7923,
+      "step": 888
+    },
+    {
+      "epoch": 0.2778993435448578,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019767678559159098,
+      "loss": 1.6497,
+      "step": 889
+    },
+    {
+      "epoch": 0.2782119412316349,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019767151716700845,
+      "loss": 1.9629,
+      "step": 890
+    },
+    {
+      "epoch": 0.278524538918412,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019766624284587195,
+      "loss": 1.8348,
+      "step": 891
+    },
+    {
+      "epoch": 0.2788371366051891,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019766096262849994,
+      "loss": 1.8409,
+      "step": 892
+    },
+    {
+      "epoch": 0.27914973429196627,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019765567651521115,
+      "loss": 1.7796,
+      "step": 893
+    },
+    {
+      "epoch": 0.2794623319787434,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019765038450632476,
+      "loss": 1.9009,
+      "step": 894
+    },
+    {
+      "epoch": 0.2797749296655205,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019764508660216019,
+      "loss": 1.4491,
+      "step": 895
+    },
+    {
+      "epoch": 0.2800875273522976,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001976397828030373,
+      "loss": 1.5436,
+      "step": 896
+    },
+    {
+      "epoch": 0.2804001250390747,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001976344731092763,
+      "loss": 1.8577,
+      "step": 897
+    },
+    {
+      "epoch": 0.28071272272585185,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001976291575211978,
+      "loss": 2.6341,
+      "step": 898
+    },
+    {
+      "epoch": 0.28102532041262895,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019762383603912258,
+      "loss": 1.6624,
+      "step": 899
+    },
+    {
+      "epoch": 0.28133791809940606,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.000197618508663372,
+      "loss": 1.6193,
+      "step": 900
+    },
+    {
+      "epoch": 0.28165051578618316,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019761317539426765,
+      "loss": 1.6416,
+      "step": 901
+    },
+    {
+      "epoch": 0.2819631134729603,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019760783623213153,
+      "loss": 1.5813,
+      "step": 902
+    },
+    {
+      "epoch": 0.28227571115973743,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019760249117728592,
+      "loss": 2.1245,
+      "step": 903
+    },
+    {
+      "epoch": 0.28258830884651454,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019759714023005357,
+      "loss": 2.0305,
+      "step": 904
+    },
+    {
+      "epoch": 0.28290090653329164,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001975917833907575,
+      "loss": 1.4689,
+      "step": 905
+    },
+    {
+      "epoch": 0.28321350422006875,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00019758642065972112,
+      "loss": 1.9306,
+      "step": 906
+    },
+    {
+      "epoch": 0.2835261019068459,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001975810520372681,
+      "loss": 1.8309,
+      "step": 907
+    },
+    {
+      "epoch": 0.283838699593623,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001975756775237227,
+      "loss": 1.732,
+      "step": 908
+    },
+    {
+      "epoch": 0.2841512972804001,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019757029711940923,
+      "loss": 1.5233,
+      "step": 909
+    },
+    {
+      "epoch": 0.2844638949671772,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019756491082465263,
+      "loss": 1.6491,
+      "step": 910
+    },
+    {
+      "epoch": 0.2847764926539544,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019755951863977805,
+      "loss": 2.2236,
+      "step": 911
+    },
+    {
+      "epoch": 0.2850890903407315,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019755412056511097,
+      "loss": 1.8299,
+      "step": 912
+    },
+    {
+      "epoch": 0.2854016880275086,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019754871660097734,
+      "loss": 1.5403,
+      "step": 913
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019754330674770339,
+      "loss": 1.5712,
+      "step": 914
+    },
+    {
+      "epoch": 0.28602688340106286,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019753789100561569,
+      "loss": 1.8814,
+      "step": 915
+    },
+    {
+      "epoch": 0.28633948108783996,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001975324693750412,
+      "loss": 1.7153,
+      "step": 916
+    },
+    {
+      "epoch": 0.28665207877461707,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001975270418563073,
+      "loss": 2.0221,
+      "step": 917
+    },
+    {
+      "epoch": 0.2869646764613942,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019752160844974158,
+      "loss": 1.7176,
+      "step": 918
+    },
+    {
+      "epoch": 0.2872772741481713,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001975161691556721,
+      "loss": 1.8581,
+      "step": 919
+    },
+    {
+      "epoch": 0.28758987183494844,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019751072397442715,
+      "loss": 1.8127,
+      "step": 920
+    },
+    {
+      "epoch": 0.28790246952172555,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001975052729063356,
+      "loss": 1.5183,
+      "step": 921
+    },
+    {
+      "epoch": 0.28821506720850265,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019749981595172647,
+      "loss": 1.7887,
+      "step": 922
+    },
+    {
+      "epoch": 0.28852766489527976,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019749435311092919,
+      "loss": 1.7053,
+      "step": 923
+    },
+    {
+      "epoch": 0.2888402625820569,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019748888438427358,
+      "loss": 1.7008,
+      "step": 924
+    },
+    {
+      "epoch": 0.289152860268834,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019748340977208977,
+      "loss": 1.888,
+      "step": 925
+    },
+    {
+      "epoch": 0.2894654579556111,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001974779292747083,
+      "loss": 1.708,
+      "step": 926
+    },
+    {
+      "epoch": 0.28977805564238823,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019747244289246006,
+      "loss": 1.8244,
+      "step": 927
+    },
+    {
+      "epoch": 0.29009065332916534,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001974669506256762,
+      "loss": 1.5614,
+      "step": 928
+    },
+    {
+      "epoch": 0.2904032510159425,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019746145247468832,
+      "loss": 2.2925,
+      "step": 929
+    },
+    {
+      "epoch": 0.2907158487027196,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019745594843982836,
+      "loss": 1.7933,
+      "step": 930
+    },
+    {
+      "epoch": 0.2910284463894967,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001974504385214286,
+      "loss": 1.8521,
+      "step": 931
+    },
+    {
+      "epoch": 0.2913410440762738,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019744492271982168,
+      "loss": 1.6939,
+      "step": 932
+    },
+    {
+      "epoch": 0.291653641763051,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019743940103534062,
+      "loss": 1.6783,
+      "step": 933
+    },
+    {
+      "epoch": 0.2919662394498281,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019743387346831876,
+      "loss": 2.0204,
+      "step": 934
+    },
+    {
+      "epoch": 0.2922788371366052,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019742834001908977,
+      "loss": 1.7812,
+      "step": 935
+    },
+    {
+      "epoch": 0.2925914348233823,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019742280068798775,
+      "loss": 1.7483,
+      "step": 936
+    },
+    {
+      "epoch": 0.29290403251015945,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019741725547534712,
+      "loss": 1.8223,
+      "step": 937
+    },
+    {
+      "epoch": 0.29321663019693656,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001974117043815026,
+      "loss": 1.8306,
+      "step": 938
+    },
+    {
+      "epoch": 0.29352922788371366,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019740614740678937,
+      "loss": 1.9111,
+      "step": 939
+    },
+    {
+      "epoch": 0.29384182557049077,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001974005845515429,
+      "loss": 1.7384,
+      "step": 940
+    },
+    {
+      "epoch": 0.29415442325726787,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019739501581609903,
+      "loss": 1.7809,
+      "step": 941
+    },
+    {
+      "epoch": 0.29446702094404503,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019738944120079393,
+      "loss": 1.8266,
+      "step": 942
+    },
+    {
+      "epoch": 0.29477961863082214,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001973838607059642,
+      "loss": 2.0459,
+      "step": 943
+    },
+    {
+      "epoch": 0.29509221631759924,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019737827433194665,
+      "loss": 1.7519,
+      "step": 944
+    },
+    {
+      "epoch": 0.29540481400437635,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001973726820790786,
+      "loss": 1.6264,
+      "step": 945
+    },
+    {
+      "epoch": 0.2957174116911535,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019736708394769764,
+      "loss": 1.6892,
+      "step": 946
+    },
+    {
+      "epoch": 0.2960300093779306,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019736147993814176,
+      "loss": 1.9491,
+      "step": 947
+    },
+    {
+      "epoch": 0.2963426070647077,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019735587005074927,
+      "loss": 1.7754,
+      "step": 948
+    },
+    {
+      "epoch": 0.2966552047514848,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019735025428585886,
+      "loss": 1.9126,
+      "step": 949
+    },
+    {
+      "epoch": 0.29696780243826193,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019734463264380953,
+      "loss": 2.071,
+      "step": 950
+    },
+    {
+      "epoch": 0.2972804001250391,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001973390051249407,
+      "loss": 1.6336,
+      "step": 951
+    },
+    {
+      "epoch": 0.2975929978118162,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019733337172959204,
+      "loss": 1.4598,
+      "step": 952
+    },
+    {
+      "epoch": 0.2979055954985933,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001973277324581037,
+      "loss": 1.5984,
+      "step": 953
+    },
+    {
+      "epoch": 0.2982181931853704,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019732208731081615,
+      "loss": 1.9082,
+      "step": 954
+    },
+    {
+      "epoch": 0.29853079087214757,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019731643628807018,
+      "loss": 1.6075,
+      "step": 955
+    },
+    {
+      "epoch": 0.29884338855892467,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019731077939020693,
+      "loss": 1.9933,
+      "step": 956
+    },
+    {
+      "epoch": 0.2991559862457018,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019730511661756792,
+      "loss": 1.5719,
+      "step": 957
+    },
+    {
+      "epoch": 0.2994685839324789,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019729944797049502,
+      "loss": 1.6318,
+      "step": 958
+    },
+    {
+      "epoch": 0.29978118161925604,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019729377344933043,
+      "loss": 1.8574,
+      "step": 959
+    },
+    {
+      "epoch": 0.30009377930603315,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001972880930544168,
+      "loss": 1.9144,
+      "step": 960
+    },
+    {
+      "epoch": 0.30040637699281025,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000197282406786097,
+      "loss": 1.7335,
+      "step": 961
+    },
+    {
+      "epoch": 0.30071897467958736,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019727671464471436,
+      "loss": 1.7289,
+      "step": 962
+    },
+    {
+      "epoch": 0.30103157236636446,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019727101663061247,
+      "loss": 1.994,
+      "step": 963
+    },
+    {
+      "epoch": 0.3013441700531416,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019726531274413532,
+      "loss": 1.7233,
+      "step": 964
+    },
+    {
+      "epoch": 0.30165676773991873,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019725960298562733,
+      "loss": 1.8961,
+      "step": 965
+    },
+    {
+      "epoch": 0.30196936542669583,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019725388735543318,
+      "loss": 1.6978,
+      "step": 966
+    },
+    {
+      "epoch": 0.30228196311347294,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001972481658538979,
+      "loss": 1.752,
+      "step": 967
+    },
+    {
+      "epoch": 0.3025945608002501,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019724243848136692,
+      "loss": 2.0531,
+      "step": 968
+    },
+    {
+      "epoch": 0.3029071584870272,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.000197236705238186,
+      "loss": 1.7117,
+      "step": 969
+    },
+    {
+      "epoch": 0.3032197561738043,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019723096612470133,
+      "loss": 1.5911,
+      "step": 970
+    },
+    {
+      "epoch": 0.3035323538605814,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019722522114125929,
+      "loss": 1.8811,
+      "step": 971
+    },
+    {
+      "epoch": 0.3038449515473586,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019721947028820676,
+      "loss": 1.6444,
+      "step": 972
+    },
+    {
+      "epoch": 0.3041575492341357,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001972137135658909,
+      "loss": 1.5187,
+      "step": 973
+    },
+    {
+      "epoch": 0.3044701469209128,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001972079509746593,
+      "loss": 1.6957,
+      "step": 974
+    },
+    {
+      "epoch": 0.3047827446076899,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019720218251485983,
+      "loss": 1.5887,
+      "step": 975
+    },
+    {
+      "epoch": 0.305095342294467,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001971964081868407,
+      "loss": 1.7837,
+      "step": 976
+    },
+    {
+      "epoch": 0.30540793998124416,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001971906279909506,
+      "loss": 1.8848,
+      "step": 977
+    },
+    {
+      "epoch": 0.30572053766802126,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001971848419275384,
+      "loss": 1.8966,
+      "step": 978
+    },
+    {
+      "epoch": 0.30603313535479837,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019717904999695348,
+      "loss": 1.6581,
+      "step": 979
+    },
+    {
+      "epoch": 0.3063457330415755,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019717325219954543,
+      "loss": 1.6071,
+      "step": 980
+    },
+    {
+      "epoch": 0.30665833072835263,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019716744853566436,
+      "loss": 1.8169,
+      "step": 981
+    },
+    {
+      "epoch": 0.30697092841512974,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001971616390056606,
+      "loss": 1.6017,
+      "step": 982
+    },
+    {
+      "epoch": 0.30728352610190685,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019715582360988482,
+      "loss": 1.6999,
+      "step": 983
+    },
+    {
+      "epoch": 0.30759612378868395,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019715000234868821,
+      "loss": 1.7758,
+      "step": 984
+    },
+    {
+      "epoch": 0.30790872147546106,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019714417522242214,
+      "loss": 1.9776,
+      "step": 985
+    },
+    {
+      "epoch": 0.3082213191622382,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019713834223143844,
+      "loss": 1.7776,
+      "step": 986
+    },
+    {
+      "epoch": 0.3085339168490153,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019713250337608922,
+      "loss": 1.8847,
+      "step": 987
+    },
+    {
+      "epoch": 0.3088465145357924,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.000197126658656727,
+      "loss": 1.8091,
+      "step": 988
+    },
+    {
+      "epoch": 0.30915911222256953,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019712080807370464,
+      "loss": 1.804,
+      "step": 989
+    },
+    {
+      "epoch": 0.3094717099093467,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019711495162737529,
+      "loss": 1.782,
+      "step": 990
+    },
+    {
+      "epoch": 0.3097843075961238,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001971090893180926,
+      "loss": 1.5211,
+      "step": 991
+    },
+    {
+      "epoch": 0.3100969052829009,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001971032211462104,
+      "loss": 1.4168,
+      "step": 992
+    },
+    {
+      "epoch": 0.310409502969678,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019709734711208303,
+      "loss": 1.5656,
+      "step": 993
+    },
+    {
+      "epoch": 0.31072210065645517,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019709146721606509,
+      "loss": 1.818,
+      "step": 994
+    },
+    {
+      "epoch": 0.3110346983432323,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019708558145851152,
+      "loss": 1.7158,
+      "step": 995
+    },
+    {
+      "epoch": 0.3113472960300094,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001970796898397777,
+      "loss": 1.6944,
+      "step": 996
+    },
+    {
+      "epoch": 0.3116598937167865,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001970737923602193,
+      "loss": 1.7961,
+      "step": 997
+    },
+    {
+      "epoch": 0.3119724914035636,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019706788902019233,
+      "loss": 1.8871,
+      "step": 998
+    },
+    {
+      "epoch": 0.31228508909034075,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019706197982005322,
+      "loss": 1.8513,
+      "step": 999
+    },
+    {
+      "epoch": 0.31259768677711786,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001970560647601587,
+      "loss": 1.6529,
+      "step": 1000
+    },
+    {
+      "epoch": 0.31291028446389496,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001970501438408659,
+      "loss": 1.7564,
+      "step": 1001
+    },
+    {
+      "epoch": 0.31322288215067207,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001970442170625322,
+      "loss": 1.5718,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3135354798374492,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019703828442551547,
+      "loss": 1.9791,
+      "step": 1003
+    },
+    {
+      "epoch": 0.31384807752422633,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019703234593017386,
+      "loss": 1.5583,
+      "step": 1004
+    },
+    {
+      "epoch": 0.31416067521100344,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019702640157686586,
+      "loss": 1.8005,
+      "step": 1005
+    },
+    {
+      "epoch": 0.31447327289778054,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019702045136595032,
+      "loss": 2.0622,
+      "step": 1006
+    },
+    {
+      "epoch": 0.31478587058455765,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019701449529778656,
+      "loss": 1.6313,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3150984682713348,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019700853337273406,
+      "loss": 1.7088,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3154110659581119,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001970025655911528,
+      "loss": 1.7942,
+      "step": 1009
+    },
+    {
+      "epoch": 0.315723663644889,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019699659195340303,
+      "loss": 1.8139,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3160362613316661,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001969906124598454,
+      "loss": 1.6704,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3163488590184433,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019698462711084091,
+      "loss": 1.9731,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3166614567052204,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019697863590675086,
+      "loss": 1.6923,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3169740543919975,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019697263884793702,
+      "loss": 1.8974,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3172866520787746,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001969666359347614,
+      "loss": 2.0298,
+      "step": 1015
+    },
+    {
+      "epoch": 0.31759924976555176,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019696062716758638,
+      "loss": 1.6155,
+      "step": 1016
+    },
+    {
+      "epoch": 0.31791184745232887,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019695461254677475,
+      "loss": 1.6622,
+      "step": 1017
+    },
+    {
+      "epoch": 0.31822444513910597,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019694859207268958,
+      "loss": 2.0245,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3185370428258831,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001969425657456944,
+      "loss": 1.7654,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3188496405126602,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019693653356615297,
+      "loss": 1.6629,
+      "step": 1020
+    },
+    {
+      "epoch": 0.31916223819943734,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019693049553442952,
+      "loss": 1.7823,
+      "step": 1021
+    },
+    {
+      "epoch": 0.31947483588621445,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001969244516508885,
+      "loss": 1.5993,
+      "step": 1022
+    },
+    {
+      "epoch": 0.31978743357299155,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001969184019158948,
+      "loss": 1.7385,
+      "step": 1023
+    },
+    {
+      "epoch": 0.32010003125976866,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019691234632981372,
+      "loss": 2.0781,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3204126289465458,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019690628489301077,
+      "loss": 1.6396,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3207252266333229,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019690021760585192,
+      "loss": 1.7066,
+      "step": 1026
+    },
+    {
+      "epoch": 0.32103782432010003,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019689414446870344,
+      "loss": 1.6741,
+      "step": 1027
+    },
+    {
+      "epoch": 0.32135042200687713,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.000196888065481932,
+      "loss": 1.8628,
+      "step": 1028
+    },
+    {
+      "epoch": 0.32166301969365424,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019688198064590458,
+      "loss": 1.8129,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3219756173804314,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019687588996098853,
+      "loss": 1.9068,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3222882150672085,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019686979342755154,
+      "loss": 1.8664,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3226008127539856,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001968636910459617,
+      "loss": 1.7239,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3229134104407627,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019685758281658738,
+      "loss": 1.9294,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3232260081275399,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019685146873979736,
+      "loss": 1.7469,
+      "step": 1034
+    },
+    {
+      "epoch": 0.323538605814317,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019684534881596078,
+      "loss": 1.8425,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3238512035010941,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019683922304544705,
+      "loss": 1.5658,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3241638011878712,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000196833091428626,
+      "loss": 1.7025,
+      "step": 1037
+    },
+    {
+      "epoch": 0.32447639887464835,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019682695396586785,
+      "loss": 1.7166,
+      "step": 1038
+    },
+    {
+      "epoch": 0.32478899656142546,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019682081065754313,
+      "loss": 1.8159,
+      "step": 1039
+    },
+    {
+      "epoch": 0.32510159424820256,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019681466150402266,
+      "loss": 1.7957,
+      "step": 1040
+    },
+    {
+      "epoch": 0.32541419193497967,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001968085065056777,
+      "loss": 1.6375,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3257267896217568,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019680234566287985,
+      "loss": 2.1855,
+      "step": 1042
+    },
+    {
+      "epoch": 0.32603938730853393,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019679617897600102,
+      "loss": 1.8348,
+      "step": 1043
+    },
+    {
+      "epoch": 0.32635198499531104,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019679000644541356,
+      "loss": 1.6444,
+      "step": 1044
+    },
+    {
+      "epoch": 0.32666458268208814,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019678382807149003,
+      "loss": 1.8918,
+      "step": 1045
+    },
+    {
+      "epoch": 0.32697718036886525,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019677764385460348,
+      "loss": 1.6544,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3272897780556424,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019677145379512723,
+      "loss": 1.8734,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3276023757424195,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019676525789343502,
+      "loss": 1.8792,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3279149734291966,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00019675905614990085,
+      "loss": 1.8914,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3282275711159737,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001967528485648992,
+      "loss": 1.6186,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3285401688027509,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019674663513880475,
+      "loss": 1.7937,
+      "step": 1051
+    },
+    {
+      "epoch": 0.328852766489528,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019674041587199268,
+      "loss": 1.7155,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3291653641763051,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001967341907648384,
+      "loss": 1.8787,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3294779618630822,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019672795981771777,
+      "loss": 1.6195,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3297905595498593,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019672172303100696,
+      "loss": 1.9987,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33010315723663647,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019671548040508244,
+      "loss": 1.6107,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3304157549234136,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019670923194032116,
+      "loss": 1.6394,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3307283526101907,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019670297763710028,
+      "loss": 1.7142,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3310409502969678,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00019669671749579742,
+      "loss": 1.8344,
+      "step": 1059
+    },
+    {
+      "epoch": 0.33135354798374494,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001966904515167905,
+      "loss": 1.933,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33166614567052205,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001966841797004578,
+      "loss": 1.763,
+      "step": 1061
+    },
+    {
+      "epoch": 0.33197874335729916,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.000196677902047178,
+      "loss": 1.8741,
+      "step": 1062
+    },
+    {
+      "epoch": 0.33229134104407626,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019667161855733002,
+      "loss": 1.8624,
+      "step": 1063
+    },
+    {
+      "epoch": 0.33260393873085337,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019666532923129327,
+      "loss": 1.899,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3329165364176305,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00019665903406944737,
+      "loss": 2.3084,
+      "step": 1065
+    },
+    {
+      "epoch": 0.33322913410440763,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00019665273307217245,
+      "loss": 1.6737,
+      "step": 1066
+    },
+    {
+      "epoch": 0.33354173179118474,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019664642623984886,
+      "loss": 1.6899,
+      "step": 1067
+    },
+    {
+      "epoch": 0.33385432947796184,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019664011357285735,
+      "loss": 1.8702,
+      "step": 1068
+    },
+    {
+      "epoch": 0.334166927164739,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019663379507157903,
+      "loss": 2.0766,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3344795248515161,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00019662747073639537,
+      "loss": 1.9336,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3347921225382932,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019662114056768815,
+      "loss": 1.8872,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3351047202250703,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019661480456583958,
+      "loss": 1.7719,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3354173179118475,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019660846273123213,
+      "loss": 1.695,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3357299155986246,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019660211506424867,
+      "loss": 1.8269,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3360425132854017,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001965957615652724,
+      "loss": 1.8746,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3363551109721788,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019658940223468693,
+      "loss": 1.5041,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3366677086589559,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019658303707287617,
+      "loss": 1.8079,
+      "step": 1077
+    },
+    {
+      "epoch": 0.33698030634573306,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019657666608022438,
+      "loss": 1.7644,
+      "step": 1078
+    },
+    {
+      "epoch": 0.33729290403251017,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019657028925711617,
+      "loss": 1.759,
+      "step": 1079
+    },
+    {
+      "epoch": 0.33760550171928727,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019656390660393659,
+      "loss": 1.9192,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3379180994060644,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019655751812107085,
+      "loss": 1.9153,
+      "step": 1081
+    },
+    {
+      "epoch": 0.33823069709284154,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019655112380890475,
+      "loss": 1.688,
+      "step": 1082
+    },
+    {
+      "epoch": 0.33854329477961864,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019654472366782425,
+      "loss": 1.907,
+      "step": 1083
+    },
+    {
+      "epoch": 0.33885589246639575,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019653831769821575,
+      "loss": 1.9453,
+      "step": 1084
+    },
+    {
+      "epoch": 0.33916849015317285,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.000196531905900466,
+      "loss": 1.6311,
+      "step": 1085
+    },
+    {
+      "epoch": 0.33948108783994996,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019652548827496207,
+      "loss": 1.9493,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3397936855267271,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001965190648220914,
+      "loss": 1.8175,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3401062832135042,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001965126355422418,
+      "loss": 1.8018,
+      "step": 1088
+    },
+    {
+      "epoch": 0.34041888090028133,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001965062004358014,
+      "loss": 1.6674,
+      "step": 1089
+    },
+    {
+      "epoch": 0.34073147858705843,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001964997595031587,
+      "loss": 1.9538,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3410440762738356,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019649331274470256,
+      "loss": 1.8417,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3413566739606127,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019648686016082216,
+      "loss": 2.0019,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3416692716473898,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019648040175190707,
+      "loss": 1.7955,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3419818693341669,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019647393751834718,
+      "loss": 1.6747,
+      "step": 1094
+    },
+    {
+      "epoch": 0.34229446702094407,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019646746746053274,
+      "loss": 1.7818,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3426070647077212,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019646099157885437,
+      "loss": 1.7983,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3429196623944983,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019645450987370298,
+      "loss": 1.677,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3432322600812754,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019644802234546993,
+      "loss": 1.9241,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3435448577680525,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001964415289945469,
+      "loss": 1.9008,
+      "step": 1099
+    },
+    {
+      "epoch": 0.34385745545482965,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019643502982132581,
+      "loss": 1.6438,
+      "step": 1100
+    },
+    {
+      "epoch": 0.34417005314160676,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001964285248261991,
+      "loss": 1.7665,
+      "step": 1101
+    },
+    {
+      "epoch": 0.34448265082838386,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001964220140095595,
+      "loss": 1.7259,
+      "step": 1102
+    },
+    {
+      "epoch": 0.34479524851516097,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019641549737180002,
+      "loss": 1.7119,
+      "step": 1103
+    },
+    {
+      "epoch": 0.34510784620193813,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019640897491331408,
+      "loss": 1.6551,
+      "step": 1104
+    },
+    {
+      "epoch": 0.34542044388871523,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001964024466344955,
+      "loss": 1.9882,
+      "step": 1105
+    },
+    {
+      "epoch": 0.34573304157549234,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019639591253573836,
+      "loss": 1.7573,
+      "step": 1106
+    },
+    {
+      "epoch": 0.34604563926226944,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019638937261743714,
+      "loss": 1.6814,
+      "step": 1107
+    },
+    {
+      "epoch": 0.34635823694904655,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019638282687998667,
+      "loss": 1.943,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3466708346358237,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019637627532378212,
+      "loss": 1.6896,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3469834323226008,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000196369717949219,
+      "loss": 1.8984,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3472960300093779,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019636315475669324,
+      "loss": 1.4845,
+      "step": 1111
+    },
+    {
+      "epoch": 0.347608627696155,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019635658574660098,
+      "loss": 1.7234,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3479212253829322,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001963500109193389,
+      "loss": 1.5583,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3482338230697093,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019634343027530383,
+      "loss": 1.8789,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3485464207564864,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019633684381489315,
+      "loss": 2.0262,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3488590184432635,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019633025153850442,
+      "loss": 1.7877,
+      "step": 1116
+    },
+    {
+      "epoch": 0.34917161613004066,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019632365344653563,
+      "loss": 1.7381,
+      "step": 1117
+    },
+    {
+      "epoch": 0.34948421381681777,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019631704953938518,
+      "loss": 1.7758,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3497968115035949,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001963104398174517,
+      "loss": 1.8063,
+      "step": 1119
+    },
+    {
+      "epoch": 0.350109409190372,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019630382428113417,
+      "loss": 1.8691,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3504220068771491,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019629720293083214,
+      "loss": 1.7844,
+      "step": 1121
+    },
+    {
+      "epoch": 0.35073460456392624,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019629057576694522,
+      "loss": 1.6097,
+      "step": 1122
+    },
+    {
+      "epoch": 0.35104720225070335,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019628394278987355,
+      "loss": 1.9393,
+      "step": 1123
+    },
+    {
+      "epoch": 0.35135979993748045,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001962773040000175,
+      "loss": 1.7556,
+      "step": 1124
+    },
+    {
+      "epoch": 0.35167239762425756,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000196270659397778,
+      "loss": 1.7145,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3519849953110347,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001962640089835561,
+      "loss": 1.6505,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3522975929978118,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019625735275775327,
+      "loss": 1.6953,
+      "step": 1127
+    },
+    {
+      "epoch": 0.35261019068458893,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019625069072077138,
+      "loss": 1.7897,
+      "step": 1128
+    },
+    {
+      "epoch": 0.35292278837136604,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001962440228730127,
+      "loss": 1.8916,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3532353860581432,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019623734921487965,
+      "loss": 1.5444,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3535479837449203,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019623066974677525,
+      "loss": 1.6391,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3538605814316974,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019622398446910263,
+      "loss": 1.6171,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3541731791184745,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001962172933822655,
+      "loss": 1.6352,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3544857768052516,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019621059648666772,
+      "loss": 1.8147,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3547983744920288,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019620389378271366,
+      "loss": 1.7773,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3551109721788059,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001961971852708079,
+      "loss": 1.7441,
+      "step": 1136
+    },
+    {
+      "epoch": 0.355423569865583,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019619047095135553,
+      "loss": 1.9931,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3557361675523601,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019618375082476182,
+      "loss": 1.6723,
+      "step": 1138
+    },
+    {
+      "epoch": 0.35604876523913725,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001961770248914325,
+      "loss": 1.8312,
+      "step": 1139
+    },
+    {
+      "epoch": 0.35636136292591436,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019617029315177365,
+      "loss": 2.0553,
+      "step": 1140
+    },
+    {
+      "epoch": 0.35667396061269147,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019616355560619163,
+      "loss": 1.6513,
+      "step": 1141
+    },
+    {
+      "epoch": 0.35698655829946857,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019615681225509325,
+      "loss": 1.8244,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3572991559862457,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019615006309888552,
+      "loss": 1.9322,
+      "step": 1143
+    },
+    {
+      "epoch": 0.35761175367302284,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000196143308137976,
+      "loss": 1.7572,
+      "step": 1144
+    },
+    {
+      "epoch": 0.35792435135979994,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019613654737277245,
+      "loss": 1.5536,
+      "step": 1145
+    },
+    {
+      "epoch": 0.35823694904657705,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000196129780803683,
+      "loss": 1.9036,
+      "step": 1146
+    },
+    {
+      "epoch": 0.35854954673335415,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019612300843111622,
+      "loss": 2.1856,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3588621444201313,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001961162302554809,
+      "loss": 1.6396,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3591747421069084,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019610944627718627,
+      "loss": 1.8837,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3594873397936855,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019610265649664193,
+      "loss": 1.7418,
+      "step": 1150
+    },
+    {
+      "epoch": 0.35979993748046263,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019609586091425774,
+      "loss": 1.8848,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3601125351672398,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019608905953044396,
+      "loss": 1.4857,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3604251328540169,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019608225234561121,
+      "loss": 1.6741,
+      "step": 1153
+    },
+    {
+      "epoch": 0.360737730540794,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019607543936017046,
+      "loss": 1.6363,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3610503282275711,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019606862057453298,
+      "loss": 1.8323,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3613629259143482,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019606179598911049,
+      "loss": 1.6778,
+      "step": 1156
+    },
+    {
+      "epoch": 0.36167552360112537,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019605496560431496,
+      "loss": 1.8691,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3619881212879025,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019604812942055873,
+      "loss": 1.6175,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3623007189746796,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019604128743825453,
+      "loss": 1.717,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3626133166614567,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019603443965781543,
+      "loss": 1.773,
+      "step": 1160
+    },
+    {
+      "epoch": 0.36292591434823385,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019602758607965484,
+      "loss": 1.8844,
+      "step": 1161
+    },
+    {
+      "epoch": 0.36323851203501095,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019602072670418647,
+      "loss": 1.9545,
+      "step": 1162
+    },
+    {
+      "epoch": 0.36355110972178806,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019601386153182451,
+      "loss": 1.523,
+      "step": 1163
+    },
+    {
+      "epoch": 0.36386370740856516,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019600699056298337,
+      "loss": 2.0468,
+      "step": 1164
+    },
+    {
+      "epoch": 0.36417630509534227,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019600011379807786,
+      "loss": 1.9032,
+      "step": 1165
+    },
+    {
+      "epoch": 0.36448890278211943,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019599323123752315,
+      "loss": 1.3631,
+      "step": 1166
+    },
+    {
+      "epoch": 0.36480150046889653,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019598634288173474,
+      "loss": 1.6805,
+      "step": 1167
+    },
+    {
+      "epoch": 0.36511409815567364,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019597944873112852,
+      "loss": 1.4813,
+      "step": 1168
+    },
+    {
+      "epoch": 0.36542669584245074,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019597254878612065,
+      "loss": 1.7945,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3657392935292279,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001959656430471277,
+      "loss": 1.5851,
+      "step": 1170
+    },
+    {
+      "epoch": 0.366051891216005,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001959587315145666,
+      "loss": 1.8493,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3663644889027821,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001959518141888546,
+      "loss": 1.7852,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3666770865895592,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019594489107040928,
+      "loss": 1.9668,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3669896842763364,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019593796215964867,
+      "loss": 1.656,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3673022819631135,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000195931027456991,
+      "loss": 1.5947,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3676148796498906,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019592408696285496,
+      "loss": 1.7685,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3679274773366677,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019591714067765953,
+      "loss": 1.6027,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3682400750234448,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001959101886018241,
+      "loss": 2.2013,
+      "step": 1178
+    },
+    {
+      "epoch": 0.36855267271022196,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001959032307357684,
+      "loss": 1.6995,
+      "step": 1179
+    },
+    {
+      "epoch": 0.36886527039699907,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019589626707991242,
+      "loss": 1.7104,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3691778680837762,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019588929763467657,
+      "loss": 1.6798,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3694904657705533,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019588232240048167,
+      "loss": 1.5464,
+      "step": 1182
+    },
+    {
+      "epoch": 0.36980306345733044,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001958753413777488,
+      "loss": 1.7789,
+      "step": 1183
+    },
+    {
+      "epoch": 0.37011566114410754,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019586835456689934,
+      "loss": 1.7634,
+      "step": 1184
+    },
+    {
+      "epoch": 0.37042825883088465,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001958613619683552,
+      "loss": 1.9015,
+      "step": 1185
+    },
+    {
+      "epoch": 0.37074085651766175,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019585436358253845,
+      "loss": 2.3964,
+      "step": 1186
+    },
+    {
+      "epoch": 0.37105345420443886,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019584735940987163,
+      "loss": 1.7068,
+      "step": 1187
+    },
+    {
+      "epoch": 0.371366051891216,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019584034945077758,
+      "loss": 1.9431,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3716786495779931,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001958333337056795,
+      "loss": 1.6602,
+      "step": 1189
+    },
+    {
+      "epoch": 0.37199124726477023,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019582631217500093,
+      "loss": 1.9655,
+      "step": 1190
+    },
+    {
+      "epoch": 0.37230384495154734,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001958192848591658,
+      "loss": 1.7755,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3726164426383245,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019581225175859833,
+      "loss": 1.7425,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3729290403251016,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019580521287372317,
+      "loss": 1.8308,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3732416380118787,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019579816820496516,
+      "loss": 1.7996,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3735542356986558,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001957911177527497,
+      "loss": 1.8265,
+      "step": 1195
+    },
+    {
+      "epoch": 0.373866833385433,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019578406151750236,
+      "loss": 1.5686,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3741794310722101,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001957769994996492,
+      "loss": 1.7951,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3744920287589872,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019576993169961653,
+      "loss": 1.7821,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3748046264457643,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.000195762858117831,
+      "loss": 1.7286,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3751172241325414,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019575577875471974,
+      "loss": 1.707,
+      "step": 1200
+    },
+    {
+      "epoch": 0.37542982181931855,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019574869361071006,
+      "loss": 1.9656,
+      "step": 1201
+    },
+    {
+      "epoch": 0.37574241950609566,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019574160268622976,
+      "loss": 1.7242,
+      "step": 1202
+    },
+    {
+      "epoch": 0.37605501719287276,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019573450598170687,
+      "loss": 1.7001,
+      "step": 1203
+    },
+    {
+      "epoch": 0.37636761487964987,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019572740349756992,
+      "loss": 1.8952,
+      "step": 1204
+    },
+    {
+      "epoch": 0.37668021256642703,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019572029523424756,
+      "loss": 1.8052,
+      "step": 1205
+    },
+    {
+      "epoch": 0.37699281025320414,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019571318119216904,
+      "loss": 1.8727,
+      "step": 1206
+    },
+    {
+      "epoch": 0.37730540793998124,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001957060613717638,
+      "loss": 1.6054,
+      "step": 1207
+    },
+    {
+      "epoch": 0.37761800562675835,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019569893577346168,
+      "loss": 1.8537,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3779306033135355,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019569180439769283,
+      "loss": 1.6096,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3782432010003126,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019568466724488782,
+      "loss": 1.9668,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3785557986870897,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019567752431547754,
+      "loss": 1.6992,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3788683963738668,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019567037560989315,
+      "loss": 1.6169,
+      "step": 1212
+    },
+    {
+      "epoch": 0.37918099406064393,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019566322112856633,
+      "loss": 1.7126,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3794935917474211,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001956560608719289,
+      "loss": 1.6279,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3798061894341982,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001956488948404132,
+      "loss": 2.0578,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3801187871209753,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019564172303445182,
+      "loss": 1.7761,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3804313848077524,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019563454545447773,
+      "loss": 1.6644,
+      "step": 1217
+    },
+    {
+      "epoch": 0.38074398249452956,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019562736210092428,
+      "loss": 1.8542,
+      "step": 1218
+    },
+    {
+      "epoch": 0.38105658018130667,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001956201729742251,
+      "loss": 1.7917,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3813691778680838,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019561297807481427,
+      "loss": 1.8474,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3816817755548609,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001956057774031261,
+      "loss": 1.627,
+      "step": 1221
+    },
+    {
+      "epoch": 0.381994373241638,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019559857095959528,
+      "loss": 1.6842,
+      "step": 1222
+    },
+    {
+      "epoch": 0.38230697092841515,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019559135874465695,
+      "loss": 1.7735,
+      "step": 1223
+    },
+    {
+      "epoch": 0.38261956861519225,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019558414075874646,
+      "loss": 1.8281,
+      "step": 1224
+    },
+    {
+      "epoch": 0.38293216630196936,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019557691700229957,
+      "loss": 1.5633,
+      "step": 1225
+    },
+    {
+      "epoch": 0.38324476398874646,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019556968747575244,
+      "loss": 1.8649,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3835573616755236,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019556245217954149,
+      "loss": 1.6938,
+      "step": 1227
+    },
+    {
+      "epoch": 0.38386995936230073,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001955552111141035,
+      "loss": 1.6866,
+      "step": 1228
+    },
+    {
+      "epoch": 0.38418255704907783,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019554796427987566,
+      "loss": 1.9343,
+      "step": 1229
+    },
+    {
+      "epoch": 0.38449515473585494,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019554071167729545,
+      "loss": 1.9785,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3848077524226321,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019553345330680077,
+      "loss": 1.876,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3851203501094092,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019552618916882973,
+      "loss": 1.671,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3854329477961863,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019551891926382093,
+      "loss": 1.6575,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3857455454829634,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019551164359221326,
+      "loss": 1.9775,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3860581431697405,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019550436215444594,
+      "loss": 1.7329,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3863707408565177,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001954970749509586,
+      "loss": 1.6745,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3866833385432948,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019548978198219113,
+      "loss": 1.7502,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3869959362300719,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019548248324858386,
+      "loss": 1.6299,
+      "step": 1238
+    },
+    {
+      "epoch": 0.387308533916849,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019547517875057738,
+      "loss": 1.6477,
+      "step": 1239
+    },
+    {
+      "epoch": 0.38762113160362616,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019546786848861268,
+      "loss": 1.8717,
+      "step": 1240
+    },
+    {
+      "epoch": 0.38793372929040326,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019546055246313113,
+      "loss": 1.5382,
+      "step": 1241
+    },
+    {
+      "epoch": 0.38824632697718037,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019545323067457439,
+      "loss": 2.0394,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3885589246639575,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019544590312338444,
+      "loss": 1.8064,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3888715223507346,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019543856981000371,
+      "loss": 1.6846,
+      "step": 1244
+    },
+    {
+      "epoch": 0.38918412003751174,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001954312307348749,
+      "loss": 1.7834,
+      "step": 1245
+    },
+    {
+      "epoch": 0.38949671772428884,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001954238858984411,
+      "loss": 1.8043,
+      "step": 1246
+    },
+    {
+      "epoch": 0.38980931541106595,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019541653530114568,
+      "loss": 1.7905,
+      "step": 1247
+    },
+    {
+      "epoch": 0.39012191309784305,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019540917894343246,
+      "loss": 1.6521,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3904345107846202,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019540181682574552,
+      "loss": 1.6881,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3907471084713973,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001953944489485293,
+      "loss": 2.0565,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3910597061581744,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019538707531222867,
+      "loss": 1.7884,
+      "step": 1251
+    },
+    {
+      "epoch": 0.39137230384495153,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019537969591728872,
+      "loss": 1.5153,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3916849015317287,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000195372310764155,
+      "loss": 1.8401,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3919974992185058,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019536491985327334,
+      "loss": 1.5898,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3923100969052829,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019535752318508998,
+      "loss": 1.8118,
+      "step": 1255
+    },
+    {
+      "epoch": 0.39262269459206,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019535012076005138,
+      "loss": 1.4033,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3929352922788371,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019534271257860448,
+      "loss": 1.672,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3932478899656143,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019533529864119658,
+      "loss": 1.752,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3935604876523914,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001953278789482752,
+      "loss": 1.3813,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3938730853391685,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019532045350028826,
+      "loss": 1.8827,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3941856830259456,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019531302229768404,
+      "loss": 1.9363,
+      "step": 1261
+    },
+    {
+      "epoch": 0.39449828071272275,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019530558534091127,
+      "loss": 1.8975,
+      "step": 1262
+    },
+    {
+      "epoch": 0.39481087839949985,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019529814263041884,
+      "loss": 1.7931,
+      "step": 1263
+    },
+    {
+      "epoch": 0.39512347608627696,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001952906941666561,
+      "loss": 1.7258,
+      "step": 1264
+    },
+    {
+      "epoch": 0.39543607377305406,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001952832399500727,
+      "loss": 1.8547,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3957486714598312,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019527577998111874,
+      "loss": 1.7344,
+      "step": 1266
+    },
+    {
+      "epoch": 0.39606126914660833,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001952683142602445,
+      "loss": 1.7313,
+      "step": 1267
+    },
+    {
+      "epoch": 0.39637386683338544,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019526084278790074,
+      "loss": 1.8261,
+      "step": 1268
+    },
+    {
+      "epoch": 0.39668646452016254,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019525336556453852,
+      "loss": 1.7306,
+      "step": 1269
+    },
+    {
+      "epoch": 0.39699906220693965,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001952458825906092,
+      "loss": 1.9536,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3973116598937168,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019523839386656458,
+      "loss": 1.7486,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3976242575804939,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019523089939285675,
+      "loss": 1.9232,
+      "step": 1272
+    },
+    {
+      "epoch": 0.397936855267271,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001952233991699382,
+      "loss": 1.5959,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3982494529540481,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019521589319826168,
+      "loss": 1.9811,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3985620506408253,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019520838147828035,
+      "loss": 1.6908,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3988746483276024,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019520086401044772,
+      "loss": 1.7011,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3991872460143795,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001951933407952176,
+      "loss": 1.6478,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3994998437011566,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001951858118330442,
+      "loss": 1.5169,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3998124413879337,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019517827712438207,
+      "loss": 1.7061,
+      "step": 1279
+    },
+    {
+      "epoch": 0.40012503907471086,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019517073666968604,
+      "loss": 1.7499,
+      "step": 1280
+    },
+    {
+      "epoch": 0.40043763676148797,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019516319046941134,
+      "loss": 2.132,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4007502344482651,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019515563852401358,
+      "loss": 1.56,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4010628321350422,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019514808083394866,
+      "loss": 1.86,
+      "step": 1283
+    },
+    {
+      "epoch": 0.40137542982181934,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019514051739967286,
+      "loss": 1.6877,
+      "step": 1284
+    },
+    {
+      "epoch": 0.40168802750859645,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019513294822164277,
+      "loss": 1.5612,
+      "step": 1285
+    },
+    {
+      "epoch": 0.40200062519537355,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019512537330031537,
+      "loss": 1.7812,
+      "step": 1286
+    },
+    {
+      "epoch": 0.40231322288215066,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019511779263614798,
+      "loss": 1.5228,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4026258205689278,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019511020622959823,
+      "loss": 1.4276,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4029384182557049,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019510261408112414,
+      "loss": 1.8561,
+      "step": 1289
+    },
+    {
+      "epoch": 0.403251015942482,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019509501619118403,
+      "loss": 1.8674,
+      "step": 1290
+    },
+    {
+      "epoch": 0.40356361362925913,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001950874125602366,
+      "loss": 1.8583,
+      "step": 1291
+    },
+    {
+      "epoch": 0.40387621131603624,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019507980318874096,
+      "loss": 1.686,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4041888090028134,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019507218807715638,
+      "loss": 1.7897,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4045014066895905,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019506456722594265,
+      "loss": 1.7626,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4048140043763676,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001950569406355599,
+      "loss": 1.9098,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4051266020631447,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001950493083064685,
+      "loss": 1.5848,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4054391997499219,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019504167023912922,
+      "loss": 1.6362,
+      "step": 1297
+    },
+    {
+      "epoch": 0.405751797436699,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001950340264340032,
+      "loss": 1.9604,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4060643951234761,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001950263768915519,
+      "loss": 2.5325,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4063769928102532,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019501872161223712,
+      "loss": 1.9979,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4066895904970303,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019501106059652108,
+      "loss": 1.714,
+      "step": 1301
+    },
+    {
+      "epoch": 0.40700218818380746,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001950033938448662,
+      "loss": 1.7827,
+      "step": 1302
+    },
+    {
+      "epoch": 0.40731478587058456,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019499572135773537,
+      "loss": 1.6062,
+      "step": 1303
+    },
+    {
+      "epoch": 0.40762738355736167,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001949880431355918,
+      "loss": 1.6599,
+      "step": 1304
+    },
+    {
+      "epoch": 0.40793998124413877,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000194980359178899,
+      "loss": 1.5345,
+      "step": 1305
+    },
+    {
+      "epoch": 0.40825257893091593,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001949726694881209,
+      "loss": 1.8149,
+      "step": 1306
+    },
+    {
+      "epoch": 0.40856517661769304,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019496497406372174,
+      "loss": 1.6207,
+      "step": 1307
+    },
+    {
+      "epoch": 0.40887777430447014,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019495727290616606,
+      "loss": 1.7058,
+      "step": 1308
+    },
+    {
+      "epoch": 0.40919037199124725,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001949495660159188,
+      "loss": 1.5045,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4095029696780244,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019494185339344523,
+      "loss": 1.8221,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4098155673648015,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.000194934135039211,
+      "loss": 1.4478,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4101281650515786,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001949264109536821,
+      "loss": 1.4922,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4104407627383557,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019491868113732474,
+      "loss": 1.8462,
+      "step": 1313
+    },
+    {
+      "epoch": 0.41075336042513283,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001949109455906057,
+      "loss": 1.831,
+      "step": 1314
+    },
+    {
+      "epoch": 0.41106595811191,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001949032043139919,
+      "loss": 1.5742,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4113785557986871,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001948954573079507,
+      "loss": 1.7099,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4116911534854642,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019488770457294985,
+      "loss": 1.8771,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4120037511722413,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019487994610945734,
+      "loss": 1.9056,
+      "step": 1318
+    },
+    {
+      "epoch": 0.41231634885901847,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019487218191794158,
+      "loss": 1.7384,
+      "step": 1319
+    },
+    {
+      "epoch": 0.41262894654579557,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019486441199887132,
+      "loss": 1.9079,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4129415442325727,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019485663635271562,
+      "loss": 1.8313,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4132541419193498,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019484885497994387,
+      "loss": 1.642,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4135667396061269,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019484106788102593,
+      "loss": 1.7165,
+      "step": 1323
+    },
+    {
+      "epoch": 0.41387933729290405,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001948332750564318,
+      "loss": 1.6474,
+      "step": 1324
+    },
+    {
+      "epoch": 0.41419193497968115,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019482547650663206,
+      "loss": 1.5541,
+      "step": 1325
+    },
+    {
+      "epoch": 0.41450453266645826,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019481767223209745,
+      "loss": 2.0118,
+      "step": 1326
+    },
+    {
+      "epoch": 0.41481713035323536,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019480986223329913,
+      "loss": 1.8306,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4151297280400125,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019480204651070864,
+      "loss": 1.6828,
+      "step": 1328
+    },
+    {
+      "epoch": 0.41544232572678963,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019479422506479775,
+      "loss": 1.6071,
+      "step": 1329
+    },
+    {
+      "epoch": 0.41575492341356673,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00019478639789603872,
+      "loss": 1.6847,
+      "step": 1330
+    },
+    {
+      "epoch": 0.41606752110034384,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019477856500490405,
+      "loss": 1.6309,
+      "step": 1331
+    },
+    {
+      "epoch": 0.416380118787121,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019477072639186664,
+      "loss": 1.9451,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4166927164738981,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001947628820573997,
+      "loss": 1.8675,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4170053141606752,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019475503200197685,
+      "loss": 1.5601,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4173179118474523,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019474717622607195,
+      "loss": 1.5294,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4176305095342294,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019473931473015926,
+      "loss": 1.7433,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4179431072210066,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019473144751471345,
+      "loss": 1.6771,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4182557049077837,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001947235745802094,
+      "loss": 1.9994,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4185683025945608,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001947156959271225,
+      "loss": 1.726,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4188809002813379,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019470781155592827,
+      "loss": 1.8079,
+      "step": 1340
+    },
+    {
+      "epoch": 0.41919349796811506,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019469992146710282,
+      "loss": 1.8046,
+      "step": 1341
+    },
+    {
+      "epoch": 0.41950609565489216,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001946920256611224,
+      "loss": 1.619,
+      "step": 1342
+    },
+    {
+      "epoch": 0.41981869334166927,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019468412413846373,
+      "loss": 1.6015,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4201312910284464,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019467621689960385,
+      "loss": 1.7538,
+      "step": 1344
+    },
+    {
+      "epoch": 0.42044388871522353,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019466830394502009,
+      "loss": 1.8732,
+      "step": 1345
+    },
+    {
+      "epoch": 0.42075648640200064,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001946603852751902,
+      "loss": 1.7492,
+      "step": 1346
+    },
+    {
+      "epoch": 0.42106908408877775,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001946524608905922,
+      "loss": 1.6893,
+      "step": 1347
+    },
+    {
+      "epoch": 0.42138168177555485,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019464453079170454,
+      "loss": 1.5848,
+      "step": 1348
+    },
+    {
+      "epoch": 0.42169427946233196,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019463659497900593,
+      "loss": 1.5974,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4220068771491091,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001946286534529755,
+      "loss": 1.9757,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4223194748358862,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001946207062140927,
+      "loss": 1.9514,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4226320725226633,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019461275326283724,
+      "loss": 1.894,
+      "step": 1352
+    },
+    {
+      "epoch": 0.42294467020944043,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019460479459968932,
+      "loss": 1.5872,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4232572678962176,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001945968302251294,
+      "loss": 1.5275,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4235698655829947,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001945888601396383,
+      "loss": 1.6427,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4238824632697718,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019458088434369715,
+      "loss": 1.6407,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4241950609565489,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019457290283778747,
+      "loss": 1.9373,
+      "step": 1357
+    },
+    {
+      "epoch": 0.424507658643326,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001945649156223912,
+      "loss": 1.7385,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4248202563301032,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001945569226979904,
+      "loss": 1.8262,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4251328540168803,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019454892406506775,
+      "loss": 1.6286,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4254454517036574,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019454091972410603,
+      "loss": 1.7992,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4257580493904345,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001945329096755885,
+      "loss": 1.9609,
+      "step": 1362
+    },
+    {
+      "epoch": 0.42607064707721165,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019452489391999874,
+      "loss": 1.9051,
+      "step": 1363
+    },
+    {
+      "epoch": 0.42638324476398876,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019451687245782072,
+      "loss": 1.7331,
+      "step": 1364
+    },
+    {
+      "epoch": 0.42669584245076586,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019450884528953864,
+      "loss": 2.1455,
+      "step": 1365
+    },
+    {
+      "epoch": 0.42700844013754297,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019450081241563716,
+      "loss": 1.8298,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4273210378243201,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019449277383660118,
+      "loss": 1.8084,
+      "step": 1367
+    },
+    {
+      "epoch": 0.42763363551109723,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019448472955291605,
+      "loss": 1.6876,
+      "step": 1368
+    },
+    {
+      "epoch": 0.42794623319787434,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001944766795650674,
+      "loss": 1.7431,
+      "step": 1369
+    },
+    {
+      "epoch": 0.42825883088465144,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001944686238735412,
+      "loss": 1.7904,
+      "step": 1370
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019446056247882378,
+      "loss": 1.8465,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4288840262582057,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019445249538140185,
+      "loss": 1.6672,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4291966239449828,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001944444225817624,
+      "loss": 1.9209,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4295092216317599,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019443634408039282,
+      "loss": 1.8336,
+      "step": 1374
+    },
+    {
+      "epoch": 0.429821819318537,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001944282598777808,
+      "loss": 1.9261,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4301344170053142,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001944201699744144,
+      "loss": 1.6371,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4304470146920913,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019441207437078203,
+      "loss": 1.4774,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4307596123788684,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001944039730673724,
+      "loss": 1.5849,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4310722100656455,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001943958660646746,
+      "loss": 1.8103,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4313848077524226,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019438775336317812,
+      "loss": 1.8946,
+      "step": 1380
+    },
+    {
+      "epoch": 0.43169740543919977,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019437963496337266,
+      "loss": 1.6056,
+      "step": 1381
+    },
+    {
+      "epoch": 0.43201000312597687,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019437151086574837,
+      "loss": 1.6991,
+      "step": 1382
+    },
+    {
+      "epoch": 0.432322600812754,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019436338107079574,
+      "loss": 1.6126,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4326351984995311,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019435524557900551,
+      "loss": 1.4967,
+      "step": 1384
+    },
+    {
+      "epoch": 0.43294779618630824,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019434710439086888,
+      "loss": 1.5868,
+      "step": 1385
+    },
+    {
+      "epoch": 0.43326039387308535,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019433895750687734,
+      "loss": 1.7528,
+      "step": 1386
+    },
+    {
+      "epoch": 0.43357299155986245,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019433080492752268,
+      "loss": 1.899,
+      "step": 1387
+    },
+    {
+      "epoch": 0.43388558924663956,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019432264665329715,
+      "loss": 2.0873,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4341981869334167,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019431448268469325,
+      "loss": 1.4453,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4345107846201938,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019430631302220385,
+      "loss": 1.9314,
+      "step": 1390
+    },
+    {
+      "epoch": 0.43482338230697093,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001942981376663221,
+      "loss": 1.5989,
+      "step": 1391
+    },
+    {
+      "epoch": 0.43513597999374803,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019428995661754171,
+      "loss": 1.8037,
+      "step": 1392
+    },
+    {
+      "epoch": 0.43544857768052514,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001942817698763564,
+      "loss": 1.7903,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4357611753673023,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019427357744326057,
+      "loss": 1.7809,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4360737730540794,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001942653793187487,
+      "loss": 1.552,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4363863707408565,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019425717550331572,
+      "loss": 1.7079,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4366989684276336,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000194248965997457,
+      "loss": 1.8321,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4370115661144108,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019424075080166805,
+      "loss": 1.6185,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4373241638011879,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019423252991644492,
+      "loss": 1.7149,
+      "step": 1399
+    },
+    {
+      "epoch": 0.437636761487965,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019422430334228386,
+      "loss": 1.7048,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4379493591747421,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019421607107968154,
+      "loss": 1.8062,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4382619568615192,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019420783312913494,
+      "loss": 1.8332,
+      "step": 1402
+    },
+    {
+      "epoch": 0.43857455454829636,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0001941995894911414,
+      "loss": 2.397,
+      "step": 1403
+    },
+    {
+      "epoch": 0.43888715223507346,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019419134016619865,
+      "loss": 1.6672,
+      "step": 1404
+    },
+    {
+      "epoch": 0.43919974992185057,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001941830851548046,
+      "loss": 1.6112,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4395123476086277,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001941748244574577,
+      "loss": 1.7182,
+      "step": 1406
+    },
+    {
+      "epoch": 0.43982494529540483,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019416655807465667,
+      "loss": 1.7438,
+      "step": 1407
+    },
+    {
+      "epoch": 0.44013754298218194,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001941582860069005,
+      "loss": 1.8327,
+      "step": 1408
+    },
+    {
+      "epoch": 0.44045014066895904,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019415000825468863,
+      "loss": 2.0563,
+      "step": 1409
+    },
+    {
+      "epoch": 0.44076273835573615,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001941417248185208,
+      "loss": 1.9451,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4410753360425133,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019413343569889702,
+      "loss": 1.8786,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4413879337292904,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019412514089631785,
+      "loss": 1.7905,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4417005314160675,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019411684041128392,
+      "loss": 1.7573,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4420131291028446,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019410853424429642,
+      "loss": 1.6898,
+      "step": 1414
+    },
+    {
+      "epoch": 0.44232572678962173,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019410022239585678,
+      "loss": 1.7676,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4426383244763989,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001940919048664668,
+      "loss": 1.7774,
+      "step": 1416
+    },
+    {
+      "epoch": 0.442950922163176,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019408358165662866,
+      "loss": 1.6328,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4432635198499531,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019407525276684474,
+      "loss": 1.7037,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4435761175367302,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019406691819761796,
+      "loss": 1.81,
+      "step": 1419
+    },
+    {
+      "epoch": 0.44388871522350737,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019405857794945147,
+      "loss": 1.8474,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4442013129102845,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019405023202284874,
+      "loss": 1.6398,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4445139105970616,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001940418804183137,
+      "loss": 1.5592,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4448265082838387,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019403352313635046,
+      "loss": 1.6566,
+      "step": 1423
+    },
+    {
+      "epoch": 0.44513910597061584,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001940251601774636,
+      "loss": 1.6928,
+      "step": 1424
+    },
+    {
+      "epoch": 0.44545170365739295,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019401679154215802,
+      "loss": 2.029,
+      "step": 1425
+    },
+    {
+      "epoch": 0.44576430134417006,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001940084172309389,
+      "loss": 1.9225,
+      "step": 1426
+    },
+    {
+      "epoch": 0.44607689903094716,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00019400003724431185,
+      "loss": 1.9033,
+      "step": 1427
+    },
+    {
+      "epoch": 0.44638949671772427,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019399165158278279,
+      "loss": 1.9373,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4467020944045014,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019398326024685792,
+      "loss": 1.8287,
+      "step": 1429
+    },
+    {
+      "epoch": 0.44701469209127853,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019397486323704388,
+      "loss": 1.4876,
+      "step": 1430
+    },
+    {
+      "epoch": 0.44732728977805564,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001939664605538476,
+      "loss": 1.7532,
+      "step": 1431
+    },
+    {
+      "epoch": 0.44763988746483274,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001939580521977763,
+      "loss": 1.8811,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4479524851516099,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019394963816933772,
+      "loss": 1.8956,
+      "step": 1433
+    },
+    {
+      "epoch": 0.448265082838387,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019394121846903975,
+      "loss": 1.7634,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4485776805251641,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001939327930973907,
+      "loss": 1.5284,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4488902782119412,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019392436205489924,
+      "loss": 1.8581,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4492028758987183,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019391592534207436,
+      "loss": 1.4981,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4495154735854955,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019390748295942535,
+      "loss": 1.6315,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4498280712722726,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019389903490746194,
+      "loss": 1.755,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4501406689590497,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019389058118669418,
+      "loss": 1.6564,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4504532666458268,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019388212179763235,
+      "loss": 1.8079,
+      "step": 1441
+    },
+    {
+      "epoch": 0.45076586433260396,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001938736567407872,
+      "loss": 1.7621,
+      "step": 1442
+    },
+    {
+      "epoch": 0.45107846201938107,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019386518601666977,
+      "loss": 2.0246,
+      "step": 1443
+    },
+    {
+      "epoch": 0.45139105970615817,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001938567096257914,
+      "loss": 1.7006,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4517036573929353,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019384822756866394,
+      "loss": 1.7433,
+      "step": 1445
+    },
+    {
+      "epoch": 0.45201625507971244,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019383973984579936,
+      "loss": 1.6673,
+      "step": 1446
+    },
+    {
+      "epoch": 0.45232885276648954,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019383124645771008,
+      "loss": 1.7402,
+      "step": 1447
+    },
+    {
+      "epoch": 0.45264145045326665,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019382274740490892,
+      "loss": 1.7445,
+      "step": 1448
+    },
+    {
+      "epoch": 0.45295404814004375,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001938142426879089,
+      "loss": 1.752,
+      "step": 1449
+    },
+    {
+      "epoch": 0.45326664582682086,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019380573230722353,
+      "loss": 1.7653,
+      "step": 1450
+    },
+    {
+      "epoch": 0.453579243513598,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019379721626336656,
+      "loss": 1.4672,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4538918412003751,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001937886945568521,
+      "loss": 1.6907,
+      "step": 1452
+    },
+    {
+      "epoch": 0.45420443888715223,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019378016718819466,
+      "loss": 1.7775,
+      "step": 1453
+    },
+    {
+      "epoch": 0.45451703657392933,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019377163415790902,
+      "loss": 1.913,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4548296342607065,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019376309546651033,
+      "loss": 1.8471,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4551422319474836,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019375455111451405,
+      "loss": 1.5682,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4554548296342607,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019374600110243608,
+      "loss": 1.7008,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4557674273210378,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019373744543079257,
+      "loss": 1.7075,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4560800250078149,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001937288841001,
+      "loss": 1.6143,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4563926226945921,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019372031711087527,
+      "loss": 1.6665,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4567052203813692,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019371174446363557,
+      "loss": 1.6533,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4570178180681463,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019370316615889842,
+      "loss": 1.5501,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4573304157549234,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019369458219718175,
+      "loss": 1.8101,
+      "step": 1463
+    },
+    {
+      "epoch": 0.45764301344170055,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019368599257900372,
+      "loss": 1.6708,
+      "step": 1464
+    },
+    {
+      "epoch": 0.45795561112847766,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019367739730488296,
+      "loss": 1.6922,
+      "step": 1465
+    },
+    {
+      "epoch": 0.45826820881525476,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019366879637533834,
+      "loss": 1.6808,
+      "step": 1466
+    },
+    {
+      "epoch": 0.45858080650203187,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019366018979088913,
+      "loss": 1.654,
+      "step": 1467
+    },
+    {
+      "epoch": 0.45889340418880903,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001936515775520549,
+      "loss": 1.7892,
+      "step": 1468
+    },
+    {
+      "epoch": 0.45920600187558613,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019364295965935562,
+      "loss": 1.6039,
+      "step": 1469
+    },
+    {
+      "epoch": 0.45951859956236324,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001936343361133115,
+      "loss": 1.6348,
+      "step": 1470
+    },
+    {
+      "epoch": 0.45983119724914034,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001936257069144432,
+      "loss": 2.0579,
+      "step": 1471
+    },
+    {
+      "epoch": 0.46014379493591745,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019361707206327168,
+      "loss": 1.5824,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4604563926226946,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001936084315603182,
+      "loss": 1.6563,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4607689903094717,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001935997854061044,
+      "loss": 1.7782,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4610815879962488,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019359113360115234,
+      "loss": 1.7625,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4613941856830259,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019358247614598427,
+      "loss": 1.5607,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4617067833698031,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019357381304112281,
+      "loss": 1.6091,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4620193810565802,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019356514428709104,
+      "loss": 1.5822,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4623319787433573,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001935564698844123,
+      "loss": 1.8785,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4626445764301344,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001935477898336102,
+      "loss": 1.4933,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4629571741169115,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019353910413520887,
+      "loss": 2.2543,
+      "step": 1481
+    },
+    {
+      "epoch": 0.46326977180368867,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001935304127897326,
+      "loss": 1.6022,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4635823694904658,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019352171579770615,
+      "loss": 1.9542,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4638949671772429,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019351301315965452,
+      "loss": 1.5863,
+      "step": 1484
+    },
+    {
+      "epoch": 0.46420756486402,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019350430487610312,
+      "loss": 1.9259,
+      "step": 1485
+    },
+    {
+      "epoch": 0.46452016255079714,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001934955909475777,
+      "loss": 1.9044,
+      "step": 1486
+    },
+    {
+      "epoch": 0.46483276023757425,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019348687137460432,
+      "loss": 1.829,
+      "step": 1487
+    },
+    {
+      "epoch": 0.46514535792435135,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019347814615770933,
+      "loss": 1.5524,
+      "step": 1488
+    },
+    {
+      "epoch": 0.46545795561112846,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019346941529741954,
+      "loss": 1.683,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4657705532979056,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001934606787942621,
+      "loss": 1.8919,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4660831509846827,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019345193664876433,
+      "loss": 1.7553,
+      "step": 1491
+    },
+    {
+      "epoch": 0.46639574867145983,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001934431888614541,
+      "loss": 1.9543,
+      "step": 1492
+    },
+    {
+      "epoch": 0.46670834635823694,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019343443543285945,
+      "loss": 1.6919,
+      "step": 1493
+    },
+    {
+      "epoch": 0.46702094404501404,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019342567636350887,
+      "loss": 1.6121,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4673335417317912,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019341691165393116,
+      "loss": 1.5772,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4676461394185683,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019340814130465548,
+      "loss": 1.9449,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4679587371053454,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019339936531621122,
+      "loss": 1.7063,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4682713347921225,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001933905836891283,
+      "loss": 1.7768,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4685839324788997,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019338179642393685,
+      "loss": 1.7279,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4688965301656768,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001933730035211673,
+      "loss": 1.7344,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4692091278524539,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019336420498135057,
+      "loss": 1.6349,
+      "step": 1501
+    },
+    {
+      "epoch": 0.469521725539231,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001933554008050178,
+      "loss": 1.703,
+      "step": 1502
+    },
+    {
+      "epoch": 0.46983432322600815,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019334659099270053,
+      "loss": 1.6039,
+      "step": 1503
+    },
+    {
+      "epoch": 0.47014692091278526,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001933377755449306,
+      "loss": 1.7018,
+      "step": 1504
+    },
+    {
+      "epoch": 0.47045951859956237,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019332895446224022,
+      "loss": 1.5957,
+      "step": 1505
+    },
+    {
+      "epoch": 0.47077211628633947,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019332012774516191,
+      "loss": 1.6054,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4710847139731166,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001933112953942286,
+      "loss": 1.6822,
+      "step": 1507
+    },
+    {
+      "epoch": 0.47139731165989374,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019330245740997346,
+      "loss": 1.6045,
+      "step": 1508
+    },
+    {
+      "epoch": 0.47170990934667084,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019329361379293006,
+      "loss": 1.6817,
+      "step": 1509
+    },
+    {
+      "epoch": 0.47202250703344795,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019328476454363237,
+      "loss": 1.6334,
+      "step": 1510
+    },
+    {
+      "epoch": 0.47233510472022505,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019327590966261452,
+      "loss": 1.9416,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4726477024070022,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019326704915041115,
+      "loss": 1.8148,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4729603000937793,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0001932581830075572,
+      "loss": 1.6804,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4732728977805564,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019324931123458784,
+      "loss": 1.6578,
+      "step": 1514
+    },
+    {
+      "epoch": 0.47358549546733353,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019324043383203875,
+      "loss": 1.7513,
+      "step": 1515
+    },
+    {
+      "epoch": 0.47389809315411063,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019323155080044587,
+      "loss": 1.8009,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4742106908408878,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019322266214034546,
+      "loss": 1.5399,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4745232885276649,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019321376785227416,
+      "loss": 1.6751,
+      "step": 1518
+    },
+    {
+      "epoch": 0.474835886214442,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019320486793676889,
+      "loss": 1.5572,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4751484839012191,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019319596239436698,
+      "loss": 1.6178,
+      "step": 1520
+    },
+    {
+      "epoch": 0.47546108158799627,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019318705122560602,
+      "loss": 1.5581,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4757736792747734,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019317813443102408,
+      "loss": 1.6904,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4760862769615505,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001931692120111594,
+      "loss": 1.9162,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4763988746483276,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001931602839665507,
+      "loss": 1.6703,
+      "step": 1524
+    },
+    {
+      "epoch": 0.47671147233510475,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001931513502977369,
+      "loss": 1.6865,
+      "step": 1525
+    },
+    {
+      "epoch": 0.47702407002188185,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019314241100525738,
+      "loss": 1.7221,
+      "step": 1526
+    },
+    {
+      "epoch": 0.47733666770865896,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019313346608965183,
+      "loss": 1.6306,
+      "step": 1527
+    },
+    {
+      "epoch": 0.47764926539543606,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019312451555146022,
+      "loss": 2.0435,
+      "step": 1528
+    },
+    {
+      "epoch": 0.47796186308221317,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019311555939122298,
+      "loss": 1.4892,
+      "step": 1529
+    },
+    {
+      "epoch": 0.47827446076899033,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019310659760948075,
+      "loss": 1.7291,
+      "step": 1530
+    },
+    {
+      "epoch": 0.47858705845576743,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019309763020677458,
+      "loss": 1.7014,
+      "step": 1531
+    },
+    {
+      "epoch": 0.47889965614254454,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019308865718364583,
+      "loss": 2.0065,
+      "step": 1532
+    },
+    {
+      "epoch": 0.47921225382932164,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019307967854063622,
+      "loss": 1.5883,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4795248515160988,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001930706942782878,
+      "loss": 1.7971,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4798374492028759,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019306170439714298,
+      "loss": 1.6701,
+      "step": 1535
+    },
+    {
+      "epoch": 0.480150046889653,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019305270889774444,
+      "loss": 1.611,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4804626445764301,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019304370778063534,
+      "loss": 1.8515,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4807752422632072,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019303470104635898,
+      "loss": 1.64,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4810878399499844,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001930256886954592,
+      "loss": 1.7283,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4814004376367615,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00019301667072848004,
+      "loss": 1.8076,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4817130353235386,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019300764714596594,
+      "loss": 1.9384,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4820256330103157,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019299861794846166,
+      "loss": 1.8492,
+      "step": 1542
+    },
+    {
+      "epoch": 0.48233823069709286,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019298958313651227,
+      "loss": 1.744,
+      "step": 1543
+    },
+    {
+      "epoch": 0.48265082838386997,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001929805427106633,
+      "loss": 1.7691,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4829634260706471,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019297149667146045,
+      "loss": 1.6095,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4832760237574242,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001929624450194499,
+      "loss": 1.8153,
+      "step": 1546
+    },
+    {
+      "epoch": 0.48358862144420134,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019295338775517803,
+      "loss": 1.8315,
+      "step": 1547
+    },
+    {
+      "epoch": 0.48390121913097844,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019294432487919173,
+      "loss": 1.6651,
+      "step": 1548
+    },
+    {
+      "epoch": 0.48421381681775555,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001929352563920381,
+      "loss": 1.632,
+      "step": 1549
+    },
+    {
+      "epoch": 0.48452641450453265,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001929261822942646,
+      "loss": 1.5682,
+      "step": 1550
+    },
+    {
+      "epoch": 0.48483901219130976,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019291710258641907,
+      "loss": 1.7631,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4851516098780869,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019290801726904962,
+      "loss": 1.6418,
+      "step": 1552
+    },
+    {
+      "epoch": 0.485464207564864,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001928989263427048,
+      "loss": 1.4744,
+      "step": 1553
+    },
+    {
+      "epoch": 0.48577680525164113,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001928898298079334,
+      "loss": 1.7507,
+      "step": 1554
+    },
+    {
+      "epoch": 0.48608940293841824,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019288072766528462,
+      "loss": 1.5483,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4864020006251954,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019287161991530792,
+      "loss": 1.7318,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4867145983119725,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001928625065585532,
+      "loss": 1.8483,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4870271959987496,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019285338759557065,
+      "loss": 1.6431,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4873397936855267,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019284426302691073,
+      "loss": 1.6648,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4876523913723038,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019283513285312437,
+      "loss": 1.5061,
+      "step": 1560
+    },
+    {
+      "epoch": 0.487964989059081,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0001928259970747627,
+      "loss": 2.72,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4882775867458581,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019281685569237734,
+      "loss": 1.6893,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4885901844326352,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001928077087065201,
+      "loss": 1.6951,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4889027821194123,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001927985561177432,
+      "loss": 1.7366,
+      "step": 1564
+    },
+    {
+      "epoch": 0.48921537980618945,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019278939792659924,
+      "loss": 1.7637,
+      "step": 1565
+    },
+    {
+      "epoch": 0.48952797749296656,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019278023413364106,
+      "loss": 1.5522,
+      "step": 1566
+    },
+    {
+      "epoch": 0.48984057517974366,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019277106473942194,
+      "loss": 1.8184,
+      "step": 1567
+    },
+    {
+      "epoch": 0.49015317286652077,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019276188974449543,
+      "loss": 1.5573,
+      "step": 1568
+    },
+    {
+      "epoch": 0.49046577055329793,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019275270914941538,
+      "loss": 1.5074,
+      "step": 1569
+    },
+    {
+      "epoch": 0.49077836824007504,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019274352295473612,
+      "loss": 1.9685,
+      "step": 1570
+    },
+    {
+      "epoch": 0.49109096592685214,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019273433116101217,
+      "loss": 1.8918,
+      "step": 1571
+    },
+    {
+      "epoch": 0.49140356361362925,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019272513376879854,
+      "loss": 1.8173,
+      "step": 1572
+    },
+    {
+      "epoch": 0.49171616130040635,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019271593077865035,
+      "loss": 1.7093,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4920287589871835,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019270672219112332,
+      "loss": 1.7993,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4923413566739606,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019269750800677331,
+      "loss": 1.7468,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4926539543607377,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00019268828822615661,
+      "loss": 1.4455,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4929665520475148,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019267906284982985,
+      "loss": 1.9409,
+      "step": 1577
+    },
+    {
+      "epoch": 0.493279149734292,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019266983187834995,
+      "loss": 1.8848,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4935917474210691,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001926605953122742,
+      "loss": 1.5927,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4939043451078462,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019265135315216028,
+      "loss": 1.7506,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4942169427946233,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019264210539856607,
+      "loss": 1.7024,
+      "step": 1581
+    },
+    {
+      "epoch": 0.49452954048140046,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001926328520520499,
+      "loss": 1.8899,
+      "step": 1582
+    },
+    {
+      "epoch": 0.49484213816817757,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001926235931131704,
+      "loss": 1.7209,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4951547358549547,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019261432858248657,
+      "loss": 1.582,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4954673335417318,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001926050584605577,
+      "loss": 1.7583,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4957799312285089,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019259578274794344,
+      "loss": 1.7366,
+      "step": 1586
+    },
+    {
+      "epoch": 0.49609252891528605,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001925865014452038,
+      "loss": 1.7721,
+      "step": 1587
+    },
+    {
+      "epoch": 0.49640512660206315,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019257721455289906,
+      "loss": 1.9818,
+      "step": 1588
+    },
+    {
+      "epoch": 0.49671772428884026,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019256792207158991,
+      "loss": 1.719,
+      "step": 1589
+    },
+    {
+      "epoch": 0.49703032197561736,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019255862400183733,
+      "loss": 1.7085,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4973429196623945,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019254932034420266,
+      "loss": 1.5593,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4976555173491716,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019254001109924763,
+      "loss": 1.6743,
+      "step": 1592
+    },
+    {
+      "epoch": 0.49796811503594873,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001925306962675342,
+      "loss": 1.5977,
+      "step": 1593
+    },
+    {
+      "epoch": 0.49828071272272584,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019252137584962472,
+      "loss": 1.6007,
+      "step": 1594
+    },
+    {
+      "epoch": 0.49859331040950294,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019251204984608184,
+      "loss": 1.5078,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4989059080962801,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019250271825746866,
+      "loss": 1.9624,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4992185057830572,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0001924933810843485,
+      "loss": 1.6749,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4995311034698343,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019248403832728504,
+      "loss": 1.7965,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4998437011566114,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019247468998684233,
+      "loss": 1.7333,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5001562988433885,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019246533606358476,
+      "loss": 1.9014,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5001562988433885,
+      "eval_loss": 1.6468836069107056,
+      "eval_runtime": 1904.4552,
+      "eval_samples_per_second": 4.798,
+      "eval_steps_per_second": 2.399,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5004688965301657,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000192455976558077,
+      "loss": 1.8399,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5007814942169428,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019244661147088413,
+      "loss": 1.7516,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5010940919037199,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019243724080257154,
+      "loss": 1.6023,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5014066895904971,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001924278645537049,
+      "loss": 1.8678,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5017192872772741,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001924184827248503,
+      "loss": 1.8877,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5020318849640513,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019240909531657415,
+      "loss": 1.7109,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5023444826508284,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019239970232944314,
+      "loss": 1.9394,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5026570803376055,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019239030376402437,
+      "loss": 1.6907,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5029696780243826,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019238089962088522,
+      "loss": 1.3726,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5032822757111597,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019237148990059342,
+      "loss": 1.4186,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5035948733979368,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019236207460371707,
+      "loss": 1.8961,
+      "step": 1611
+    },
+    {
+      "epoch": 0.503907471084714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001923526537308246,
+      "loss": 1.5122,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5042200687714911,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019234322728248473,
+      "loss": 1.6718,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5045326664582682,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019233379525926652,
+      "loss": 1.5157,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5048452641450454,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019232435766173946,
+      "loss": 1.8013,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5051578618318224,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019231491449047327,
+      "loss": 1.6126,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5054704595185996,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019230546574603805,
+      "loss": 1.9199,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5057830572053766,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019229601142900426,
+      "loss": 1.8629,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5060956548921538,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001922865515399426,
+      "loss": 1.9572,
+      "step": 1619
+    },
+    {
+      "epoch": 0.506408252578931,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0001922770860794243,
+      "loss": 1.8666,
+      "step": 1620
+    },
+    {
+      "epoch": 0.506720850265708,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019226761504802066,
+      "loss": 1.6269,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5070334479524852,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019225813844630355,
+      "loss": 1.4542,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5073460456392622,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019224865627484502,
+      "loss": 1.726,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5076586433260394,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019223916853421756,
+      "loss": 1.9227,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5079712410128165,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.000192229675224994,
+      "loss": 1.7876,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5082838386995936,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001922201763477474,
+      "loss": 1.9213,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5085964363863708,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019221067190305121,
+      "loss": 1.8536,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5089090340731478,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019220116189147928,
+      "loss": 1.7391,
+      "step": 1628
+    },
+    {
+      "epoch": 0.509221631759925,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019219164631360572,
+      "loss": 1.5871,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5095342294467021,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019218212517000497,
+      "loss": 1.7358,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5098468271334792,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019217259846125186,
+      "loss": 1.7538,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5101594248202563,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019216306618792151,
+      "loss": 2.0148,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5104720225070335,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019215352835058944,
+      "loss": 1.655,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5107846201938105,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001921439849498314,
+      "loss": 1.8552,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5110972178805877,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001921344359862236,
+      "loss": 2.0283,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5114098155673648,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019212488146034247,
+      "loss": 1.8859,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5117224132541419,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019211532137276485,
+      "loss": 1.7173,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5120350109409191,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0001921057557240679,
+      "loss": 1.6262,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5123476086276961,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019209618451482911,
+      "loss": 1.6141,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5126602063144733,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001920866077456263,
+      "loss": 1.7475,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5129728040012503,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001920770254170376,
+      "loss": 1.7333,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5132854016880275,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001920674375296416,
+      "loss": 1.8058,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5135979993748047,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019205784408401705,
+      "loss": 1.5659,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5139105970615817,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00019204824508074314,
+      "loss": 1.6922,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5142231947483589,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019203864052039937,
+      "loss": 1.5329,
+      "step": 1645
+    },
+    {
+      "epoch": 0.514535792435136,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019202903040356557,
+      "loss": 1.5799,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5148483901219131,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019201941473082196,
+      "loss": 1.7131,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5151609878086902,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019200979350274898,
+      "loss": 1.668,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5154735854954673,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019200016671992755,
+      "loss": 1.8212,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5157861831822445,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019199053438293884,
+      "loss": 1.745,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5160987808690216,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001919808964923643,
+      "loss": 1.9392,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5164113785557987,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019197125304878587,
+      "loss": 1.8001,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5167239762425758,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019196160405278567,
+      "loss": 1.6449,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5170365739293529,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019195194950494623,
+      "loss": 1.7974,
+      "step": 1654
+    },
+    {
+      "epoch": 0.51734917161613,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019194228940585043,
+      "loss": 1.6213,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5176617693029072,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001919326237560815,
+      "loss": 1.7459,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5179743669896842,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019192295255622286,
+      "loss": 2.0187,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5182869646764614,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019191327580685846,
+      "loss": 1.465,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5185995623632386,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001919035935085725,
+      "loss": 1.7626,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5189121600500156,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019189390566194943,
+      "loss": 1.6333,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5192247577367928,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019188421226757423,
+      "loss": 1.6854,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5195373554235698,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019187451332603202,
+      "loss": 1.5598,
+      "step": 1662
+    },
+    {
+      "epoch": 0.519849953110347,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019186480883790836,
+      "loss": 1.7953,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5201625507971241,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019185509880378912,
+      "loss": 1.7901,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5204751484839012,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019184538322426054,
+      "loss": 1.6819,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5207877461706784,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019183566209990911,
+      "loss": 1.8034,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5211003438574554,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019182593543132174,
+      "loss": 2.0384,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5214129415442326,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00019181620321908564,
+      "loss": 1.9369,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5217255392310097,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019180646546378832,
+      "loss": 1.8764,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5220381369177868,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019179672216601773,
+      "loss": 1.6419,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5223507346045639,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019178697332636202,
+      "loss": 2.427,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5226633322913411,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019177721894540975,
+      "loss": 1.81,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5229759299781181,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001917674590237499,
+      "loss": 1.67,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5232885276648953,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019175769356197153,
+      "loss": 1.6198,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5236011253516724,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001917479225606643,
+      "loss": 1.8033,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5239137230384495,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019173814602041803,
+      "loss": 1.6005,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5242263207252267,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019172836394182303,
+      "loss": 1.6983,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5245389184120037,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019171857632546978,
+      "loss": 1.8186,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5248515160987809,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019170878317194924,
+      "loss": 1.6052,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5251641137855579,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019169898448185256,
+      "loss": 1.7156,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5254767114723351,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019168918025577134,
+      "loss": 1.7039,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5257893091591123,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019167937049429745,
+      "loss": 1.8326,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5261019068458893,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019166955519802316,
+      "loss": 1.6872,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5264145045326665,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019165973436754098,
+      "loss": 1.6172,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5267271022194435,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019164990800344387,
+      "loss": 1.7482,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5270396999062207,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.000191640076106325,
+      "loss": 1.6177,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5273522975929978,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019163023867677797,
+      "loss": 1.6793,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5276648952797749,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019162039571539666,
+      "loss": 1.6634,
+      "step": 1688
+    },
+    {
+      "epoch": 0.527977492966552,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001916105472227753,
+      "loss": 1.7808,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5282900906533292,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019160069319950845,
+      "loss": 1.7203,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5286026883401063,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019159083364619103,
+      "loss": 1.6893,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5289152860268834,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001915809685634183,
+      "loss": 2.3232,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5292278837136605,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0001915710979517858,
+      "loss": 1.554,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5295404814004376,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001915612218118894,
+      "loss": 1.6621,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5298530790872148,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019155134014432534,
+      "loss": 1.8881,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5301656767739918,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019154145294969022,
+      "loss": 1.8313,
+      "step": 1696
+    },
+    {
+      "epoch": 0.530478274460769,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019153156022858094,
+      "loss": 1.7908,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5307908721475461,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019152166198159476,
+      "loss": 1.6425,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5311034698343232,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019151175820932917,
+      "loss": 1.7114,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5314160675211004,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00019150184891238216,
+      "loss": 1.5121,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5317286652078774,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019149193409135192,
+      "loss": 1.7762,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5320412628946546,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019148201374683704,
+      "loss": 1.8021,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5323538605814317,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019147208787943638,
+      "loss": 1.8559,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5326664582682088,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019146215648974924,
+      "loss": 2.3382,
+      "step": 1704
+    },
+    {
+      "epoch": 0.532979055954986,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019145221957837515,
+      "loss": 1.6269,
+      "step": 1705
+    },
+    {
+      "epoch": 0.533291653641763,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019144227714591402,
+      "loss": 1.8329,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5336042513285402,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001914323291929661,
+      "loss": 1.7395,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5339168490153173,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019142237572013197,
+      "loss": 1.4983,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5342294467020944,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019141241672801247,
+      "loss": 1.7625,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5345420443888715,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001914024522172089,
+      "loss": 1.8429,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5348546420756486,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019139248218832285,
+      "loss": 1.9247,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5351672397624258,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019138250664195615,
+      "loss": 1.6563,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5354798374492029,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001913725255787111,
+      "loss": 1.5108,
+      "step": 1713
+    },
+    {
+      "epoch": 0.53579243513598,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019136253899919024,
+      "loss": 1.8109,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5361050328227571,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019135254690399648,
+      "loss": 1.7063,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5364176305095343,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019134254929373303,
+      "loss": 1.7218,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5367302281963113,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019133254616900347,
+      "loss": 1.6555,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5370428258830885,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019132253753041174,
+      "loss": 1.9246,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5373554235698655,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019131252337856205,
+      "loss": 1.818,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5376680212566427,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019130250371405895,
+      "loss": 1.6691,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5379806189434199,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019129247853750733,
+      "loss": 1.6272,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5382932166301969,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001912824478495125,
+      "loss": 1.529,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5386058143169741,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019127241165067994,
+      "loss": 1.8957,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5389184120037511,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019126236994161558,
+      "loss": 1.6643,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5392310096905283,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019125232272292563,
+      "loss": 1.8746,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5395436073773054,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019124226999521672,
+      "loss": 1.5691,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5398562050640825,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019123221175909567,
+      "loss": 1.7902,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5401688027508597,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019122214801516973,
+      "loss": 1.6767,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5404814004376368,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019121207876404648,
+      "loss": 1.727,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5407939981244139,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001912020040063338,
+      "loss": 1.6355,
+      "step": 1730
+    },
+    {
+      "epoch": 0.541106595811191,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019119192374263992,
+      "loss": 1.9062,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5414191934979681,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019118183797357338,
+      "loss": 1.5986,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5417317911847452,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00019117174669974312,
+      "loss": 1.5961,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5420443888715224,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019116164992175828,
+      "loss": 1.8585,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5423569865582994,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019115154764022852,
+      "loss": 1.731,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5426695842450766,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019114143985576366,
+      "loss": 1.9891,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5429821819318537,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001911313265689739,
+      "loss": 1.6551,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5432947796186308,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00019112120778046987,
+      "loss": 2.0219,
+      "step": 1738
+    },
+    {
+      "epoch": 0.543607377305408,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001911110834908624,
+      "loss": 1.7808,
+      "step": 1739
+    },
+    {
+      "epoch": 0.543919974992185,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001911009537007627,
+      "loss": 1.7043,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5442325726789622,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019109081841078233,
+      "loss": 1.7296,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5445451703657392,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001910806776215332,
+      "loss": 1.6465,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5448577680525164,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019107053133362749,
+      "loss": 1.8411,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5451703657392936,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019106037954767774,
+      "loss": 1.4522,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5454829634260706,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019105022226429682,
+      "loss": 1.7463,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5457955611128478,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019104005948409797,
+      "loss": 1.622,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5461081587996249,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019102989120769475,
+      "loss": 1.8334,
+      "step": 1747
+    },
+    {
+      "epoch": 0.546420756486402,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019101971743570094,
+      "loss": 1.6375,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5467333541731791,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019100953816873084,
+      "loss": 1.4945,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5470459518599562,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00019099935340739893,
+      "loss": 1.687,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5473585495467334,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001909891631523201,
+      "loss": 1.8769,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5476711472335105,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019097896740410955,
+      "loss": 1.814,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5479837449202876,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019096876616338278,
+      "loss": 1.8215,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5482963426070647,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00019095855943075568,
+      "loss": 1.6682,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5486089402938418,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019094834720684447,
+      "loss": 1.8052,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5489215379806189,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001909381294922656,
+      "loss": 1.7685,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5492341356673961,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000190927906287636,
+      "loss": 1.6704,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5495467333541731,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001909176775935728,
+      "loss": 1.75,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5498593310409503,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00019090744341069356,
+      "loss": 1.5139,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5501719287277275,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019089720373961612,
+      "loss": 1.5844,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5504845264145045,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019088695858095864,
+      "loss": 1.7899,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5507971241012817,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019087670793533967,
+      "loss": 1.7717,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5511097217880587,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019086645180337803,
+      "loss": 1.7754,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5514223194748359,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001908561901856929,
+      "loss": 1.8412,
+      "step": 1764
+    },
+    {
+      "epoch": 0.551734917161613,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001908459230829038,
+      "loss": 1.7254,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5520475148483901,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019083565049563057,
+      "loss": 1.8097,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5523601125351673,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019082537242449333,
+      "loss": 1.8441,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5526727102219443,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019081508887011263,
+      "loss": 2.4757,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5529853079087215,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001908047998331093,
+      "loss": 1.5833,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5532979055954986,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001907945053141045,
+      "loss": 2.4293,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5536105032822757,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001907842053137197,
+      "loss": 1.9397,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5539231009690528,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001907738998325767,
+      "loss": 2.0662,
+      "step": 1772
+    },
+    {
+      "epoch": 0.55423569865583,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019076358887129774,
+      "loss": 1.8447,
+      "step": 1773
+    },
+    {
+      "epoch": 0.554548296342607,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00019075327243050526,
+      "loss": 2.3451,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5548608940293842,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019074295051082205,
+      "loss": 1.623,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5551734917161613,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001907326231128713,
+      "loss": 2.0579,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5554860894029384,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019072229023727645,
+      "loss": 1.6111,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5557986870897156,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019071195188466135,
+      "loss": 1.87,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5561112847764926,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00019070160805565012,
+      "loss": 1.6437,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5564238824632698,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019069125875086722,
+      "loss": 1.6752,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5567364801500468,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019068090397093745,
+      "loss": 1.7323,
+      "step": 1781
+    },
+    {
+      "epoch": 0.557049077836824,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.000190670543716486,
+      "loss": 1.7324,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5573616755236012,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019066017798813825,
+      "loss": 1.5224,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5576742732103782,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019064980678652,
+      "loss": 2.3167,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5579868708971554,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00019063943011225743,
+      "loss": 1.7731,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5582994685839325,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00019062904796597697,
+      "loss": 1.6789,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5586120662707096,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019061866034830534,
+      "loss": 1.7119,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5589246639574867,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019060826725986977,
+      "loss": 1.6962,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5592372616442638,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019059786870129761,
+      "loss": 1.6318,
+      "step": 1789
+    },
+    {
+      "epoch": 0.559549859331041,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001905874646732167,
+      "loss": 1.8541,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5598624570178181,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019057705517625505,
+      "loss": 1.8081,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5601750547045952,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001905666402110412,
+      "loss": 1.4779,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5604876523913723,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019055621977820387,
+      "loss": 1.6657,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5608002500781494,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019054579387837214,
+      "loss": 1.5665,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5611128477649265,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00019053536251217545,
+      "loss": 1.5586,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5614254454517037,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00019052492568024355,
+      "loss": 1.5323,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5617380431384807,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019051448338320656,
+      "loss": 1.7868,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5620506408252579,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019050403562169486,
+      "loss": 1.6351,
+      "step": 1798
+    },
+    {
+      "epoch": 0.562363238512035,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019049358239633916,
+      "loss": 2.0889,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5626758361988121,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00019048312370777062,
+      "loss": 1.5398,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5629884338855893,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019047265955662054,
+      "loss": 1.6967,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5633010315723663,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00019046218994352076,
+      "loss": 1.6917,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5636136292591435,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001904517148691033,
+      "loss": 1.4587,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5639262269459207,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019044123433400052,
+      "loss": 1.8214,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5642388246326977,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001904307483388452,
+      "loss": 1.6375,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5645514223194749,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019042025688427035,
+      "loss": 1.5963,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5648640200062519,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019040975997090936,
+      "loss": 1.8623,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5651766176930291,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00019039925759939597,
+      "loss": 1.6458,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5654892153798062,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001903887497703642,
+      "loss": 1.6367,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5658018130665833,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019037823648444842,
+      "loss": 1.6211,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5661144107533604,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001903677177422833,
+      "loss": 1.5955,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5664270084401375,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00019035719354450393,
+      "loss": 1.6509,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5667396061269147,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019034666389174568,
+      "loss": 1.5193,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5670522038136918,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019033612878464412,
+      "loss": 1.8779,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5673648015004689,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019032558822383542,
+      "loss": 1.746,
+      "step": 1815
+    },
+    {
+      "epoch": 0.567677399187246,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001903150422099558,
+      "loss": 1.6802,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5679899968740232,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019030449074364204,
+      "loss": 1.8168,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5683025945608002,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019029393382553108,
+      "loss": 1.6261,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5686151922475774,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00019028337145626028,
+      "loss": 1.6126,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5689277899343544,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00019027280363646728,
+      "loss": 1.7607,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5692403876211316,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001902622303667901,
+      "loss": 1.6267,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5695529853079088,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019025165164786705,
+      "loss": 1.7209,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5698655829946858,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019024106748033679,
+      "loss": 1.4932,
+      "step": 1823
+    },
+    {
+      "epoch": 0.570178180681463,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019023047786483828,
+      "loss": 1.4764,
+      "step": 1824
+    },
+    {
+      "epoch": 0.57049077836824,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00019021988280201084,
+      "loss": 1.6664,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5708033760550172,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001902092822924941,
+      "loss": 1.5628,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5711159737417943,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00019019867633692802,
+      "loss": 1.8942,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019018806493595293,
+      "loss": 1.5664,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5717411691153486,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019017744809020942,
+      "loss": 1.4663,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5720537668021257,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00019016682580033848,
+      "loss": 1.8574,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5723663644889028,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019015619806698135,
+      "loss": 1.7824,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5726789621756799,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019014556489077965,
+      "loss": 1.5226,
+      "step": 1832
+    },
+    {
+      "epoch": 0.572991559862457,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00019013492627237532,
+      "loss": 1.8333,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5733041575492341,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00019012428221241065,
+      "loss": 1.5824,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5736167552360113,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00019011363271152822,
+      "loss": 1.7483,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5739293529227883,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00019010297777037093,
+      "loss": 1.6215,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5742419506095655,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00019009231738958206,
+      "loss": 1.6124,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5745545482963426,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00019008165156980517,
+      "loss": 1.8104,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5748671459831197,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001900709803116842,
+      "loss": 1.7839,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5751797436698969,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001900603036158634,
+      "loss": 1.6926,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5754923413566739,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00019004962148298725,
+      "loss": 1.8372,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5758049390434511,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001900389339137007,
+      "loss": 1.5496,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5761175367302281,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000190028240908649,
+      "loss": 1.7024,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5764301344170053,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00019001754246847767,
+      "loss": 1.6237,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5767427321037825,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00019000683859383258,
+      "loss": 1.6012,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5770553297905595,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00018999612928535995,
+      "loss": 1.7586,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5773679274773367,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018998541454370632,
+      "loss": 1.4823,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5776805251641138,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018997469436951854,
+      "loss": 1.5688,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5779931228508909,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0001899639687634438,
+      "loss": 2.5108,
+      "step": 1849
+    },
+    {
+      "epoch": 0.578305720537668,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018995323772612964,
+      "loss": 1.6868,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5786183182244451,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018994250125822386,
+      "loss": 1.6238,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5789309159112223,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001899317593603747,
+      "loss": 1.5826,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5792435135979994,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001899210120332306,
+      "loss": 1.6792,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5795561112847765,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018991025927744042,
+      "loss": 1.8574,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5798687089715536,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001898995010936533,
+      "loss": 1.7686,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5801813066583307,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018988873748251877,
+      "loss": 1.7198,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5804939043451078,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00018987796844468658,
+      "loss": 1.7134,
+      "step": 1857
+    },
+    {
+      "epoch": 0.580806502031885,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00018986719398080695,
+      "loss": 1.5788,
+      "step": 1858
+    },
+    {
+      "epoch": 0.581119099718662,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018985641409153026,
+      "loss": 1.6557,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5814316974054392,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018984562877750737,
+      "loss": 1.719,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5817442950922164,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018983483803938932,
+      "loss": 1.7116,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5820568927789934,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001898240418778277,
+      "loss": 1.9006,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5823694904657706,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018981324029347416,
+      "loss": 1.3191,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5826820881525476,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018980243328698088,
+      "loss": 1.7602,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5829946858393248,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018979162085900025,
+      "loss": 2.0473,
+      "step": 1865
+    },
+    {
+      "epoch": 0.583307283526102,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018978080301018503,
+      "loss": 1.7591,
+      "step": 1866
+    },
+    {
+      "epoch": 0.583619881212879,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018976997974118836,
+      "loss": 1.9532,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5839324788996562,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001897591510526636,
+      "loss": 1.8456,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5842450765864332,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018974831694526452,
+      "loss": 1.7148,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5845576742732104,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018973747741964515,
+      "loss": 1.6221,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5848702719599875,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018972663247645994,
+      "loss": 2.0677,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5851828696467646,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018971578211636359,
+      "loss": 1.4428,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5854954673335417,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018970492634001114,
+      "loss": 1.6225,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5858080650203189,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018969406514805797,
+      "loss": 1.5286,
+      "step": 1874
+    },
+    {
+      "epoch": 0.586120662707096,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018968319854115978,
+      "loss": 1.7499,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5864332603938731,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018967232651997265,
+      "loss": 1.9038,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5867458580806502,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018966144908515284,
+      "loss": 1.5464,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5870584557674273,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018965056623735713,
+      "loss": 1.6405,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5873710534542045,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00018963967797724248,
+      "loss": 1.727,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5876836511409815,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018962878430546626,
+      "loss": 1.7438,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5879962488277587,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001896178852226861,
+      "loss": 1.6973,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5883088465145357,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018960698072956,
+      "loss": 1.7813,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5886214442013129,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018959607082674632,
+      "loss": 1.8691,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5889340418880901,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018958515551490364,
+      "loss": 1.8186,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5892466395748671,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018957423479469096,
+      "loss": 1.6628,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5895592372616443,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001895633086667676,
+      "loss": 1.8004,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5898718349484214,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018955237713179314,
+      "loss": 1.781,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5901844326351985,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018954144019042759,
+      "loss": 1.7539,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5904970303219756,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018953049784333116,
+      "loss": 1.6668,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5908096280087527,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018951955009116449,
+      "loss": 1.954,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5911222256955299,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001895085969345885,
+      "loss": 1.8232,
+      "step": 1891
+    },
+    {
+      "epoch": 0.591434823382307,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018949763837426445,
+      "loss": 1.5966,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5917474210690841,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018948667441085398,
+      "loss": 1.5623,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5920600187558612,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018947570504501888,
+      "loss": 1.689,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5923726164426383,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018946473027742146,
+      "loss": 1.6939,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5926852141294154,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018945375010872426,
+      "loss": 1.7252,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5929978118161926,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001894427645395902,
+      "loss": 1.7894,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5933104095029696,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018943177357068244,
+      "loss": 1.8643,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5936230071897468,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018942077720266454,
+      "loss": 1.6017,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5939356048765239,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001894097754362004,
+      "loss": 1.514,
+      "step": 1900
+    },
+    {
+      "epoch": 0.594248202563301,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018939876827195418,
+      "loss": 1.8716,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5945608002500782,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018938775571059039,
+      "loss": 1.8103,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5948733979368552,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018937673775277388,
+      "loss": 1.5777,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5951859956236324,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001893657143991698,
+      "loss": 1.6428,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5954985933104096,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018935468565044368,
+      "loss": 2.0165,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5958111909971866,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018934365150726133,
+      "loss": 1.5724,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5961237886839638,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018933261197028885,
+      "loss": 1.9301,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5964363863707408,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001893215670401928,
+      "loss": 1.6571,
+      "step": 1908
+    },
+    {
+      "epoch": 0.596748984057518,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018931051671763988,
+      "loss": 1.7479,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5970615817442951,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018929946100329725,
+      "loss": 1.6891,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5973741794310722,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001892883998978324,
+      "loss": 1.646,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5976867771178493,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018927733340191308,
+      "loss": 1.6963,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5979993748046264,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018926626151620732,
+      "loss": 1.9789,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5983119724914036,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018925518424138361,
+      "loss": 1.9244,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5986245701781807,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018924410157811073,
+      "loss": 1.5019,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5989371678649578,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001892330135270577,
+      "loss": 1.7337,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5992497655517349,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001892219200888939,
+      "loss": 1.6027,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5995623632385121,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018921082126428912,
+      "loss": 1.6431,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5998749609252891,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018919971705391335,
+      "loss": 1.822,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6001875586120663,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018918860745843703,
+      "loss": 1.656,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6005001562988433,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018917749247853078,
+      "loss": 1.6685,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6008127539856205,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001891663721148657,
+      "loss": 1.8229,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6011253516723977,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001891552463681131,
+      "loss": 1.7224,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6014379493591747,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018914411523894467,
+      "loss": 1.9986,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6017505470459519,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001891329787280324,
+      "loss": 1.4848,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6020631447327289,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018912183683604864,
+      "loss": 1.7737,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6023757424195061,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018911068956366597,
+      "loss": 1.7155,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6026883401062832,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018909953691155745,
+      "loss": 1.7669,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6030009377930603,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018908837888039637,
+      "loss": 1.8628,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6033135354798375,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001890772154708563,
+      "loss": 1.7606,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6036261331666146,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001890660466836112,
+      "loss": 1.5453,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6039387308533917,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018905487251933542,
+      "loss": 1.7034,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6042513285401688,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018904369297870349,
+      "loss": 1.6582,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6045639262269459,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001890325080623903,
+      "loss": 1.5893,
+      "step": 1934
+    },
+    {
+      "epoch": 0.604876523913723,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018902131777107117,
+      "loss": 1.602,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6051891216005002,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018901012210542165,
+      "loss": 1.636,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6055017192872773,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018899892106611762,
+      "loss": 1.8495,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6058143169740544,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018898771465383532,
+      "loss": 1.7294,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6061269146608315,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001889765028692513,
+      "loss": 1.6063,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6064395123476086,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001889652857130424,
+      "loss": 1.5972,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6067521100343858,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018895406318588585,
+      "loss": 1.9705,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6070647077211628,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018894283528845914,
+      "loss": 1.9463,
+      "step": 1942
+    },
+    {
+      "epoch": 0.60737730540794,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018893160202144012,
+      "loss": 1.7365,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6076899030947172,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018892036338550696,
+      "loss": 1.6313,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6080025007814942,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018890911938133814,
+      "loss": 1.7297,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6083150984682714,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001888978700096125,
+      "loss": 1.5932,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6086276961550484,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018888661527100914,
+      "loss": 1.7416,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6089402938418256,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0001888753551662076,
+      "loss": 1.5615,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6092528915286027,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018886408969588756,
+      "loss": 1.9525,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6095654892153798,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001888528188607292,
+      "loss": 1.4709,
+      "step": 1950
+    },
+    {
+      "epoch": 0.609878086902157,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018884154266141296,
+      "loss": 1.6341,
+      "step": 1951
+    },
+    {
+      "epoch": 0.610190684588934,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018883026109861955,
+      "loss": 1.6915,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6105032822757112,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001888189741730301,
+      "loss": 1.7387,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6108158799624883,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000188807681885326,
+      "loss": 1.4454,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6111284776492654,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018879638423618893,
+      "loss": 1.644,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6114410753360425,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018878508122630106,
+      "loss": 1.6955,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6117536730228196,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018877377285634464,
+      "loss": 1.5826,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6120662707095967,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018876245912700243,
+      "loss": 1.7957,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6123788683963739,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018875114003895748,
+      "loss": 1.5181,
+      "step": 1959
+    },
+    {
+      "epoch": 0.612691466083151,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018873981559289308,
+      "loss": 1.7115,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6130040637699281,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018872848578949296,
+      "loss": 1.9347,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6133166614567053,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018871715062944108,
+      "loss": 1.7506,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6136292591434823,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018870581011342174,
+      "loss": 2.3271,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6139418568302595,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018869446424211962,
+      "loss": 2.0109,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6142544545170365,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018868311301621968,
+      "loss": 1.5306,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6145670522038137,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018867175643640717,
+      "loss": 1.7745,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6148796498905909,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018866039450336777,
+      "loss": 1.7684,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6151922475773679,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018864902721778734,
+      "loss": 1.738,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6155048452641451,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018863765458035218,
+      "loss": 1.6707,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6158174429509221,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018862627659174886,
+      "loss": 1.5577,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6161300406376993,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018861489325266425,
+      "loss": 1.6428,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6164426383244764,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018860350456378566,
+      "loss": 1.5885,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6167552360112535,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018859211052580057,
+      "loss": 1.3899,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6170678336980306,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001885807111393969,
+      "loss": 1.8002,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6173804313848078,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001885693064052628,
+      "loss": 1.7554,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6176930290715849,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001885578963240868,
+      "loss": 1.5717,
+      "step": 1976
+    },
+    {
+      "epoch": 0.618005626758362,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018854648089655776,
+      "loss": 1.6693,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6183182244451391,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018853506012336482,
+      "loss": 1.8787,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6186308221319162,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018852363400519745,
+      "loss": 1.6435,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6189434198186934,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018851220254274554,
+      "loss": 1.7522,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6192560175054704,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018850076573669915,
+      "loss": 1.5828,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6195686151922476,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001884893235877488,
+      "loss": 1.457,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6198812128790246,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018847787609658516,
+      "loss": 1.5991,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6201938105658018,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001884664232638994,
+      "loss": 1.598,
+      "step": 1984
+    },
+    {
+      "epoch": 0.620506408252579,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018845496509038294,
+      "loss": 1.6774,
+      "step": 1985
+    },
+    {
+      "epoch": 0.620819005939356,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018844350157672755,
+      "loss": 1.7232,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6211316036261332,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018843203272362523,
+      "loss": 1.7184,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6214442013129103,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018842055853176838,
+      "loss": 1.6561,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6217567989996874,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001884090790018498,
+      "loss": 1.5792,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6220693966864645,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001883975941345624,
+      "loss": 1.9449,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6223819943732416,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018838610393059964,
+      "loss": 2.1031,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6226945920600188,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018837460839065515,
+      "loss": 1.9063,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6230071897467959,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001883631075154229,
+      "loss": 2.1289,
+      "step": 1993
+    },
+    {
+      "epoch": 0.623319787433573,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001883516013055973,
+      "loss": 2.0025,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6236323851203501,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001883400897618729,
+      "loss": 1.8512,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6239449828071272,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001883285728849447,
+      "loss": 1.8326,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6242575804939043,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018831705067550805,
+      "loss": 1.6852,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6245701781806815,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00018830552313425845,
+      "loss": 1.8256,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6248827758674586,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001882939902618919,
+      "loss": 1.6083,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6251953735542357,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018828245205910465,
+      "loss": 1.7561,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6255079712410128,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001882709085265933,
+      "loss": 1.7635,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6258205689277899,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001882593596650547,
+      "loss": 1.8553,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6261331666145671,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001882478054751861,
+      "loss": 1.6012,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6264457643013441,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018823624595768498,
+      "loss": 1.8742,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6267583619881213,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001882246811132493,
+      "loss": 1.2608,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6270709596748985,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018821311094257716,
+      "loss": 1.5808,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6273835573616755,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018820153544636713,
+      "loss": 1.6451,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6276961550484527,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000188189954625318,
+      "loss": 1.6479,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6280087527352297,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001881783684801289,
+      "loss": 1.6755,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6283213504220069,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018816677701149939,
+      "loss": 1.6337,
+      "step": 2010
+    },
+    {
+      "epoch": 0.628633948108784,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018815518022012915,
+      "loss": 1.648,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6289465457955611,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018814357810671833,
+      "loss": 1.586,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6292591434823382,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001881319706719674,
+      "loss": 1.5722,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6295717411691153,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001881203579165771,
+      "loss": 1.946,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6298843388558925,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001881087398412485,
+      "loss": 1.7869,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6301969365426696,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.000188097116446683,
+      "loss": 1.7194,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6305095342294467,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001880854877335823,
+      "loss": 2.0099,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6308221319162238,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018807385370264848,
+      "loss": 1.8415,
+      "step": 2018
+    },
+    {
+      "epoch": 0.631134729603001,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018806221435458388,
+      "loss": 1.6398,
+      "step": 2019
+    },
+    {
+      "epoch": 0.631447327289778,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018805056969009115,
+      "loss": 1.8436,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6317599249765552,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018803891970987333,
+      "loss": 1.5016,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6320725226633322,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018802726441463375,
+      "loss": 1.5147,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6323851203501094,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018801560380507604,
+      "loss": 1.5146,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6326977180368866,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018800393788190415,
+      "loss": 1.8504,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6330103157236636,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018799226664582245,
+      "loss": 1.6024,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6333229134104408,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018798059009753542,
+      "loss": 1.8456,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6336355110972178,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018796890823774806,
+      "loss": 1.5829,
+      "step": 2027
+    },
+    {
+      "epoch": 0.633948108783995,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018795722106716562,
+      "loss": 1.8332,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6342607064707722,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018794552858649366,
+      "loss": 1.8867,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6345733041575492,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018793383079643804,
+      "loss": 1.7046,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6348859018443264,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018792212769770507,
+      "loss": 1.4539,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6351984995311035,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018791041929100115,
+      "loss": 1.7966,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6355110972178806,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001878987055770332,
+      "loss": 1.7888,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6358236949046577,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001878869865565084,
+      "loss": 1.5578,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6361362925914348,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001878752622301342,
+      "loss": 1.7211,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6364488902782119,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018786353259861847,
+      "loss": 1.5837,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6367614879649891,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001878517976626693,
+      "loss": 1.6654,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6370740856517662,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018784005742299514,
+      "loss": 1.9085,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6373866833385433,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001878283118803048,
+      "loss": 1.6215,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6376992810253204,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018781656103530737,
+      "loss": 1.9168,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6380118787120975,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001878048048887122,
+      "loss": 1.8944,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6383244763988747,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018779304344122908,
+      "loss": 1.7528,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6386370740856517,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018778127669356805,
+      "loss": 1.8204,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6389496717724289,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001877695046464395,
+      "loss": 1.7069,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6392622694592061,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001877577273005541,
+      "loss": 1.3533,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6395748671459831,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018774594465662288,
+      "loss": 1.6023,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6398874648327603,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018773415671535714,
+      "loss": 1.9426,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6402000625195373,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018772236347746856,
+      "loss": 1.7982,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6405126602063145,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018771056494366913,
+      "loss": 1.7041,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6408252578930916,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018769876111467113,
+      "loss": 1.7406,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6411378555798687,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018768695199118717,
+      "loss": 1.6077,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6414504532666458,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018767513757393016,
+      "loss": 1.7813,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6417630509534229,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018766331786361338,
+      "loss": 1.6976,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6420756486402001,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018765149286095037,
+      "loss": 1.6368,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6423882463269772,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018763966256665505,
+      "loss": 1.6045,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6427008440137543,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018762782698144163,
+      "loss": 1.5185,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6430134417005314,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018761598610602463,
+      "loss": 1.5806,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6433260393873085,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001876041399411189,
+      "loss": 1.6609,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6436386370740856,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001875922884874396,
+      "loss": 1.6643,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6439512347608628,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018758043174570222,
+      "loss": 1.5697,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6442638324476399,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018756856971662258,
+      "loss": 1.6761,
+      "step": 2061
+    },
+    {
+      "epoch": 0.644576430134417,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018755670240091677,
+      "loss": 1.5763,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6448890278211942,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001875448297993013,
+      "loss": 1.7233,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6452016255079712,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018753295191249286,
+      "loss": 1.623,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6455142231947484,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018752106874120862,
+      "loss": 1.5065,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6458268208815254,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001875091802861659,
+      "loss": 2.0689,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6461394185683026,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018749728654808242,
+      "loss": 1.7316,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6464520162550798,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001874853875276763,
+      "loss": 1.7759,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6467646139418568,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018747348322566582,
+      "loss": 1.6177,
+      "step": 2069
+    },
+    {
+      "epoch": 0.647077211628634,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001874615736427697,
+      "loss": 1.8813,
+      "step": 2070
+    },
+    {
+      "epoch": 0.647389809315411,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018744965877970696,
+      "loss": 1.6428,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6477024070021882,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018743773863719683,
+      "loss": 2.3381,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6480150046889653,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018742581321595902,
+      "loss": 1.4568,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6483276023757424,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018741388251671345,
+      "loss": 1.5651,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6486402000625195,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001874019465401804,
+      "loss": 1.8459,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6489527977492967,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018739000528708046,
+      "loss": 1.6691,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6492653954360738,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018737805875813454,
+      "loss": 1.8378,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6495779931228509,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018736610695406386,
+      "loss": 1.8245,
+      "step": 2078
+    },
+    {
+      "epoch": 0.649890590809628,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018735414987559,
+      "loss": 1.7107,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6502031884964051,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018734218752343478,
+      "loss": 1.7694,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6505157861831823,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018733021989832035,
+      "loss": 1.7134,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6508283838699593,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00018731824700096933,
+      "loss": 1.8064,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6511409815567365,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018730626883210443,
+      "loss": 1.694,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6514535792435135,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018729428539244884,
+      "loss": 1.7573,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6517661769302907,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018728229668272598,
+      "loss": 1.6263,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6520787746170679,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018727030270365965,
+      "loss": 1.846,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6523913723038449,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018725830345597396,
+      "loss": 1.7912,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6527039699906221,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001872462989403933,
+      "loss": 1.777,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6530165676773992,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018723428915764237,
+      "loss": 1.675,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6533291653641763,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018722227410844625,
+      "loss": 1.5869,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6536417630509535,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018721025379353026,
+      "loss": 1.8295,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6539543607377305,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018719822821362017,
+      "loss": 1.6437,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6542669584245077,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001871861973694419,
+      "loss": 1.8373,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6545795561112848,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018717416126172177,
+      "loss": 1.3641,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6548921537980619,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018716211989118646,
+      "loss": 1.7446,
+      "step": 2095
+    },
+    {
+      "epoch": 0.655204751484839,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018715007325856292,
+      "loss": 1.7373,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6555173491716161,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018713802136457837,
+      "loss": 1.6263,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6558299468583932,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018712596420996045,
+      "loss": 1.7508,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6561425445451704,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018711390179543703,
+      "loss": 1.8481,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6564551422319475,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018710183412173635,
+      "loss": 1.7739,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6567677399187246,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018708976118958693,
+      "loss": 1.989,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6570803376055018,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001870776829997177,
+      "loss": 1.8054,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6573929352922788,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018706559955285773,
+      "loss": 1.665,
+      "step": 2103
+    },
+    {
+      "epoch": 0.657705532979056,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001870535108497366,
+      "loss": 1.703,
+      "step": 2104
+    },
+    {
+      "epoch": 0.658018130665833,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001870414168910841,
+      "loss": 1.7818,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6583307283526102,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018702931767763028,
+      "loss": 1.5893,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6586433260393874,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001870172132101057,
+      "loss": 1.6743,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6589559237261644,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018700510348924106,
+      "loss": 1.5062,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6592685214129416,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018699298851576743,
+      "loss": 1.4517,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6595811190997186,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018698086829041627,
+      "loss": 1.7555,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6598937167864958,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001869687428139192,
+      "loss": 1.7701,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6602063144732729,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018695661208700836,
+      "loss": 1.5693,
+      "step": 2112
+    },
+    {
+      "epoch": 0.66051891216005,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000186944476110416,
+      "loss": 1.473,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6608315098468271,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018693233488487483,
+      "loss": 1.4396,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6611441075336042,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018692018841111782,
+      "loss": 1.9964,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6614567052203814,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018690803668987827,
+      "loss": 1.6639,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6617693029071585,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001868958797218898,
+      "loss": 1.7607,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6620819005939356,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018688371750788635,
+      "loss": 1.5137,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6623944982807127,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018687155004860215,
+      "loss": 1.5756,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6627070959674899,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018685937734477177,
+      "loss": 1.7926,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6630196936542669,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001868471993971301,
+      "loss": 1.7269,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6633322913410441,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001868350162064123,
+      "loss": 1.6515,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6636448890278211,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018682282777335397,
+      "loss": 1.5462,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6639574867145983,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018681063409869085,
+      "loss": 1.7719,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6642700844013755,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018679843518315913,
+      "loss": 1.9495,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6645826820881525,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001867862310274953,
+      "loss": 1.5323,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6648952797749297,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018677402163243606,
+      "loss": 1.5997,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6652078774617067,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001867618069987186,
+      "loss": 1.891,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6655204751484839,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018674958712708027,
+      "loss": 1.7805,
+      "step": 2129
+    },
+    {
+      "epoch": 0.665833072835261,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018673736201825882,
+      "loss": 1.7896,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6661456705220381,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001867251316729923,
+      "loss": 1.8483,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6664582682088153,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018671289609201907,
+      "loss": 1.8642,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6667708658955924,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001867006552760778,
+      "loss": 1.4944,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6670834635823695,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018668840922590746,
+      "loss": 1.4096,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6673960612691466,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018667615794224743,
+      "loss": 1.8447,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6677086589559237,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018666390142583724,
+      "loss": 1.7672,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6680212566427008,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018665163967741694,
+      "loss": 1.4677,
+      "step": 2137
+    },
+    {
+      "epoch": 0.668333854329478,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001866393726977267,
+      "loss": 1.9113,
+      "step": 2138
+    },
+    {
+      "epoch": 0.668646452016255,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018662710048750712,
+      "loss": 1.6074,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6689590497030322,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018661482304749915,
+      "loss": 1.9865,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6692716473898093,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 1.5433,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6695842450765864,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018659025248108288,
+      "loss": 1.7213,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6698968427633636,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018657795935615802,
+      "loss": 1.7668,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6702094404501406,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018656566100441144,
+      "loss": 1.7344,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6705220381369178,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018655335742658556,
+      "loss": 1.6451,
+      "step": 2145
+    },
+    {
+      "epoch": 0.670834635823695,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018654104862342324,
+      "loss": 1.6888,
+      "step": 2146
+    },
+    {
+      "epoch": 0.671147233510472,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018652873459566749,
+      "loss": 1.426,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6714598311972492,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018651641534406178,
+      "loss": 1.6177,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6717724288840262,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018650409086934985,
+      "loss": 1.6962,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6720850265708034,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001864917611722757,
+      "loss": 1.6879,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6723976242575805,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001864794262535837,
+      "loss": 1.9992,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6727102219443576,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001864670861140186,
+      "loss": 1.9401,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6730228196311347,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018645474075432524,
+      "loss": 1.8057,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6733354173179118,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018644239017524906,
+      "loss": 2.0631,
+      "step": 2154
+    },
+    {
+      "epoch": 0.673648015004689,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018643003437753558,
+      "loss": 1.6794,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6739606126914661,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018641767336193086,
+      "loss": 1.7738,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6742732103782432,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000186405307129181,
+      "loss": 1.8517,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6745858080650203,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018639293568003268,
+      "loss": 1.5776,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6748984057517975,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018638055901523277,
+      "loss": 1.6955,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6752110034385745,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018636817713552837,
+      "loss": 1.6111,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6755236011253517,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018635579004166712,
+      "loss": 1.8155,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6758361988121288,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00018634339773439674,
+      "loss": 1.6656,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6761487964989059,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001863310002144654,
+      "loss": 1.5922,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6764613941856831,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0001863185974826216,
+      "loss": 1.7238,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6767739918724601,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018630618953961408,
+      "loss": 1.6582,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6770865895592373,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001862937763861919,
+      "loss": 2.3931,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6773991872460143,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018628135802310446,
+      "loss": 1.7434,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6777117849327915,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001862689344511015,
+      "loss": 2.0366,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6780243826195687,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.000186256505670933,
+      "loss": 1.6197,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6783369803063457,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001862440716833494,
+      "loss": 1.5561,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6786495779931229,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018623163248910127,
+      "loss": 1.8304,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6789621756798999,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018621918808893958,
+      "loss": 1.3873,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6792747733666771,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018620673848361566,
+      "loss": 1.4493,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6795873710534542,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018619428367388103,
+      "loss": 1.7057,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6798999687402313,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001861818236604877,
+      "loss": 1.5443,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6802125664270084,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018616935844418785,
+      "loss": 1.651,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6805251641137856,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000186156888025734,
+      "loss": 1.7987,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6808377618005627,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018614441240587907,
+      "loss": 1.8154,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6811503594873398,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001861319315853762,
+      "loss": 1.7168,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6814629571741169,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001861194455649788,
+      "loss": 1.4816,
+      "step": 2180
+    },
+    {
+      "epoch": 0.681775554860894,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018610695434544074,
+      "loss": 1.5243,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6820881525476712,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018609445792751618,
+      "loss": 1.7344,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6824007502344482,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018608195631195939,
+      "loss": 1.8136,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6827133479212254,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018606944949952524,
+      "loss": 1.7538,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6830259456080024,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018605693749096876,
+      "loss": 1.8747,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6833385432947796,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018604442028704533,
+      "loss": 1.6926,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6836511409815568,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018603189788851055,
+      "loss": 1.7869,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6839637386683338,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018601937029612048,
+      "loss": 1.6719,
+      "step": 2188
+    },
+    {
+      "epoch": 0.684276336355111,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001860068375106314,
+      "loss": 1.7719,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6845889340418881,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018599429953279994,
+      "loss": 1.618,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6849015317286652,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018598175636338305,
+      "loss": 1.7768,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6852141294154424,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018596920800313798,
+      "loss": 1.9978,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6855267271022194,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001859566544528222,
+      "loss": 1.3867,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6858393247889966,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001859440957131937,
+      "loss": 1.5844,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6861519224757737,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018593153178501063,
+      "loss": 1.7227,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6864645201625508,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001859189626690315,
+      "loss": 1.8812,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6867771178493279,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018590638836601505,
+      "loss": 1.5477,
+      "step": 2197
+    },
+    {
+      "epoch": 0.687089715536105,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001858938088767205,
+      "loss": 1.8684,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6874023132228821,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018588122420190722,
+      "loss": 1.8864,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6877149109096593,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018586863434233504,
+      "loss": 1.7888,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6880275085964364,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018585603929876395,
+      "loss": 1.6452,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6883401062832135,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018584343907195437,
+      "loss": 1.585,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6886527039699907,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000185830833662667,
+      "loss": 1.7144,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6889653016567677,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018581822307166281,
+      "loss": 1.7379,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6892778993435449,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018580560729970313,
+      "loss": 1.777,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6895904970303219,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018579298634754962,
+      "loss": 1.902,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6899030947170991,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018578036021596415,
+      "loss": 1.6602,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6902156924038763,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018576772890570905,
+      "loss": 1.8837,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6905282900906533,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018575509241754685,
+      "loss": 1.6694,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6908408877774305,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018574245075224046,
+      "loss": 1.7201,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6911534854642075,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018572980391055305,
+      "loss": 1.4998,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6914660831509847,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018571715189324813,
+      "loss": 1.4607,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6917786808377618,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00018570449470108952,
+      "loss": 1.8028,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6920912785245389,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018569183233484133,
+      "loss": 1.5558,
+      "step": 2214
+    },
+    {
+      "epoch": 0.692403876211316,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018567916479526804,
+      "loss": 1.5834,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6927164738980931,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001856664920831344,
+      "loss": 1.6607,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6930290715848703,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018565381419920546,
+      "loss": 1.5378,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6933416692716474,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018564113114424662,
+      "loss": 1.8949,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6936542669584245,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018562844291902353,
+      "loss": 1.9261,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6939668646452016,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018561574952430222,
+      "loss": 2.0413,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6942794623319788,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018560305096084904,
+      "loss": 1.7628,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6945920600187558,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018559034722943056,
+      "loss": 1.6226,
+      "step": 2222
+    },
+    {
+      "epoch": 0.694904657705533,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018557763833081377,
+      "loss": 1.8693,
+      "step": 2223
+    },
+    {
+      "epoch": 0.69521725539231,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001855649242657659,
+      "loss": 1.4996,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6955298530790872,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018555220503505452,
+      "loss": 2.2346,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6958424507658644,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018553948063944749,
+      "loss": 1.773,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6961550484526414,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000185526751079713,
+      "loss": 1.8362,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6964676461394186,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018551401635661958,
+      "loss": 1.6007,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6967802438261956,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018550127647093601,
+      "loss": 1.5875,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6970928415129728,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018548853142343142,
+      "loss": 1.7156,
+      "step": 2230
+    },
+    {
+      "epoch": 0.69740543919975,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018547578121487528,
+      "loss": 1.784,
+      "step": 2231
+    },
+    {
+      "epoch": 0.697718036886527,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018546302584603727,
+      "loss": 1.6756,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6980306345733042,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001854502653176875,
+      "loss": 1.8622,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6983432322600813,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0001854374996305963,
+      "loss": 1.383,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6986558299468584,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001854247287855344,
+      "loss": 1.516,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6989684276336355,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018541195278327276,
+      "loss": 1.5284,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6992810253204126,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001853991716245827,
+      "loss": 1.4208,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6995936230071897,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001853863853102358,
+      "loss": 1.8169,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6999062206939669,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000185373593841004,
+      "loss": 1.686,
+      "step": 2239
+    },
+    {
+      "epoch": 0.700218818380744,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018536079721765956,
+      "loss": 1.4067,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7005314160675211,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018534799544097505,
+      "loss": 1.7239,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7008440137542982,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018533518851172325,
+      "loss": 1.6176,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7011566114410753,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001853223764306774,
+      "loss": 1.6086,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7014692091278525,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018530955919861096,
+      "loss": 1.5131,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7017818068146295,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001852967368162977,
+      "loss": 1.685,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7020944045014067,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018528390928451173,
+      "loss": 1.8137,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7024070021881839,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018527107660402752,
+      "loss": 1.7175,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7027195998749609,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018525823877561974,
+      "loss": 1.6921,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7030321975617381,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001852453958000634,
+      "loss": 1.9215,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7033447952485151,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018523254767813393,
+      "loss": 1.5655,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7036573929352923,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00018521969441060695,
+      "loss": 1.6418,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7039699906220694,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001852068359982584,
+      "loss": 1.8771,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7042825883088465,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018519397244186458,
+      "loss": 1.7217,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7045951859956237,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001851811037422021,
+      "loss": 1.8586,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7049077836824007,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018516822990004782,
+      "loss": 1.5904,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7052203813691779,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018515535091617898,
+      "loss": 1.6428,
+      "step": 2256
+    },
+    {
+      "epoch": 0.705532979055955,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001851424667913731,
+      "loss": 1.7164,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7058455767427321,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018512957752640799,
+      "loss": 1.7193,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7061581744295092,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018511668312206177,
+      "loss": 1.5025,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7064707721162864,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018510378357911296,
+      "loss": 1.612,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7067833698030634,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018509087889834031,
+      "loss": 1.5849,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7070959674898406,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018507796908052285,
+      "loss": 1.6807,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7074085651766177,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018506505412643995,
+      "loss": 1.6728,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7077211628633948,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018505213403687137,
+      "loss": 1.7322,
+      "step": 2264
+    },
+    {
+      "epoch": 0.708033760550172,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018503920881259703,
+      "loss": 1.6204,
+      "step": 2265
+    },
+    {
+      "epoch": 0.708346358236949,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018502627845439732,
+      "loss": 1.5918,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7086589559237262,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018501334296305285,
+      "loss": 1.8249,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7089715536105032,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018500040233934454,
+      "loss": 1.974,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7092841512972804,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018498745658405356,
+      "loss": 1.6999,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7095967489840576,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018497450569796158,
+      "loss": 1.9307,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7099093466708346,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018496154968185036,
+      "loss": 1.7392,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7102219443576118,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018494858853650213,
+      "loss": 1.7068,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7105345420443888,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001849356222626994,
+      "loss": 1.8758,
+      "step": 2273
+    },
+    {
+      "epoch": 0.710847139731166,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018492265086122488,
+      "loss": 1.6345,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7111597374179431,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001849096743328617,
+      "loss": 1.7491,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7114723351047202,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001848966926783933,
+      "loss": 1.5166,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7117849327914973,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001848837058986034,
+      "loss": 1.7068,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7120975304782745,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018487071399427599,
+      "loss": 1.7652,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7124101281650516,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018485771696619542,
+      "loss": 1.7871,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7127227258518287,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018484471481514635,
+      "loss": 1.9055,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7130353235386058,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001848317075419137,
+      "loss": 1.8693,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7133479212253829,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018481869514728279,
+      "loss": 1.548,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7136605189121601,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018480567763203918,
+      "loss": 1.614,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7139731165989371,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001847926549969687,
+      "loss": 1.4828,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018477962724285763,
+      "loss": 1.8229,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7145983119724914,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018476659437049238,
+      "loss": 1.877,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7149109096592685,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018475355638065984,
+      "loss": 1.5996,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7152235073460457,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018474051327414709,
+      "loss": 1.6033,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7155361050328227,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018472746505174156,
+      "loss": 1.6509,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7158487027195999,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018471441171423103,
+      "loss": 1.8609,
+      "step": 2290
+    },
+    {
+      "epoch": 0.716161300406377,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018470135326240347,
+      "loss": 1.8864,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7164738980931541,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001846882896970473,
+      "loss": 1.5743,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7167864957799313,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018467522101895116,
+      "loss": 1.8124,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7170990934667083,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018466214722890402,
+      "loss": 1.4247,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7174116911534855,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018464906832769517,
+      "loss": 1.5627,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7177242888402626,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001846359843161142,
+      "loss": 1.8247,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7180368865270397,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.000184622895194951,
+      "loss": 1.6003,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7183494842138168,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001846098009649958,
+      "loss": 1.6546,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7186620819005939,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018459670162703905,
+      "loss": 1.8521,
+      "step": 2299
+    },
+    {
+      "epoch": 0.718974679587371,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018458359718187165,
+      "loss": 1.7397,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7192872772741482,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001845704876302847,
+      "loss": 1.7336,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7195998749609253,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018455737297306963,
+      "loss": 1.6112,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7199124726477024,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018454425321101826,
+      "loss": 1.8522,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7202250703344796,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001845311283449225,
+      "loss": 1.6348,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7205376680212566,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018451799837557485,
+      "loss": 1.7101,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7208502657080338,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018450486330376793,
+      "loss": 1.4738,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7211628633948108,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018449172313029472,
+      "loss": 1.6334,
+      "step": 2307
+    },
+    {
+      "epoch": 0.721475461081588,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018447857785594852,
+      "loss": 1.5218,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7217880587683652,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018446542748152292,
+      "loss": 1.8324,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7221006564551422,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018445227200781185,
+      "loss": 1.8051,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7224132541419194,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001844391114356095,
+      "loss": 1.65,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7227258518286964,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018442594576571035,
+      "loss": 1.8499,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7230384495154736,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001844127749989093,
+      "loss": 1.671,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7233510472022507,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0001843995991360014,
+      "loss": 1.7405,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7236636448890278,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001843864181777822,
+      "loss": 1.8025,
+      "step": 2315
+    },
+    {
+      "epoch": 0.723976242575805,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018437323212504742,
+      "loss": 1.5695,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7242888402625821,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00018436004097859308,
+      "loss": 1.2384,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7246014379493592,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018434684473921556,
+      "loss": 1.6555,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7249140356361363,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018433364340771153,
+      "loss": 1.6447,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7252266333229134,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018432043698487797,
+      "loss": 1.6859,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7255392310096905,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001843072254715122,
+      "loss": 1.7087,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7258518286964677,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001842940088684118,
+      "loss": 1.7149,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7261644263832447,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018428078717637467,
+      "loss": 1.8408,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7264770240700219,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.000184267560396199,
+      "loss": 1.7943,
+      "step": 2324
+    },
+    {
+      "epoch": 0.726789621756799,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018425432852868333,
+      "loss": 1.7252,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7271022194435761,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001842410915746265,
+      "loss": 1.6914,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7274148171303533,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001842278495348276,
+      "loss": 1.9011,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7277274148171303,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018421460241008607,
+      "loss": 1.8245,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7280400125039075,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018420135020120172,
+      "loss": 1.8638,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7283526101906845,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018418809290897455,
+      "loss": 1.7493,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7286652078774617,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001841748305342049,
+      "loss": 1.5843,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7289778055642389,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001841615630776935,
+      "loss": 1.5289,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7292904032510159,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018414829054024128,
+      "loss": 1.6851,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7296030009377931,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001841350129226495,
+      "loss": 1.3236,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7299155986245702,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018412173022571982,
+      "loss": 1.9465,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7302281963113473,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018410844245025408,
+      "loss": 1.7362,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7305407939981244,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018409514959705448,
+      "loss": 1.7688,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7308533916849015,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001840818516669235,
+      "loss": 1.658,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7311659893716786,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018406854866066403,
+      "loss": 1.6786,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7314785870584558,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018405524057907915,
+      "loss": 1.6658,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7317911847452329,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001840419274229723,
+      "loss": 1.6022,
+      "step": 2341
+    },
+    {
+      "epoch": 0.73210378243201,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018402860919314713,
+      "loss": 1.7735,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7324163801187871,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001840152858904078,
+      "loss": 1.4977,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7327289778055642,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018400195751555858,
+      "loss": 1.7735,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7330415754923414,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018398862406940412,
+      "loss": 1.5705,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7333541731791184,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018397528555274943,
+      "loss": 1.9914,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7336667708658956,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018396194196639972,
+      "loss": 1.6567,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7339793685526728,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001839485933111606,
+      "loss": 1.5779,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7342919662394498,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018393523958783788,
+      "loss": 1.6902,
+      "step": 2349
+    },
+    {
+      "epoch": 0.734604563926227,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018392188079723786,
+      "loss": 1.8415,
+      "step": 2350
+    },
+    {
+      "epoch": 0.734917161613004,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001839085169401669,
+      "loss": 1.7724,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7352297592997812,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018389514801743186,
+      "loss": 1.4619,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7355423569865583,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018388177402983984,
+      "loss": 1.7035,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7358549546733354,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018386839497819821,
+      "loss": 1.6311,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7361675523601126,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018385501086331472,
+      "loss": 1.4891,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7364801500468896,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018384162168599735,
+      "loss": 1.7706,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7367927477336668,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00018382822744705444,
+      "loss": 1.7342,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7371053454204439,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001838148281472946,
+      "loss": 1.7338,
+      "step": 2358
+    },
+    {
+      "epoch": 0.737417943107221,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001838014237875268,
+      "loss": 1.6715,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7377305407939981,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018378801436856027,
+      "loss": 1.8231,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7380431384807753,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018377459989120452,
+      "loss": 1.6681,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7383557361675523,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018376118035626942,
+      "loss": 1.6599,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7386683338543295,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00018374775576456513,
+      "loss": 1.8036,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7389809315411066,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018373432611690208,
+      "loss": 1.8082,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7392935292278837,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0001837208914140911,
+      "loss": 1.4781,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7396061269146609,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018370745165694318,
+      "loss": 1.3993,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7399187246014379,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018369400684626976,
+      "loss": 1.5936,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7402313222882151,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018368055698288248,
+      "loss": 1.4418,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7405439199749921,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018366710206759335,
+      "loss": 1.5162,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7408565176617693,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018365364210121466,
+      "loss": 1.9776,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7411691153485465,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018364017708455895,
+      "loss": 1.3729,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7414817130353235,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001836267070184392,
+      "loss": 1.6031,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7417943107221007,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001836132319036686,
+      "loss": 1.5944,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7421069084088777,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001835997517410606,
+      "loss": 1.3653,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7424195060956549,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001835862665314291,
+      "loss": 1.7253,
+      "step": 2375
+    },
+    {
+      "epoch": 0.742732103782432,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018357277627558815,
+      "loss": 1.326,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7430447014692091,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018355928097435218,
+      "loss": 1.8161,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7433572991559863,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00018354578062853595,
+      "loss": 1.8656,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7436698968427634,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001835322752389545,
+      "loss": 1.8657,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7439824945295405,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001835187648064231,
+      "loss": 1.5715,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7442950922163176,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001835052493317575,
+      "loss": 1.7185,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7446076899030947,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018349172881577356,
+      "loss": 1.7779,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7449202875898718,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018347820325928754,
+      "loss": 1.9479,
+      "step": 2383
+    },
+    {
+      "epoch": 0.745232885276649,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018346467266311604,
+      "loss": 1.7667,
+      "step": 2384
+    },
+    {
+      "epoch": 0.745545482963426,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018345113702807585,
+      "loss": 1.4014,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7458580806502032,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018343759635498422,
+      "loss": 1.8576,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7461706783369803,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018342405064465856,
+      "loss": 1.6006,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7464832760237574,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018341049989791666,
+      "loss": 1.5874,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7467958737105346,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018339694411557655,
+      "loss": 1.6729,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7471084713973116,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018338338329845668,
+      "loss": 1.5282,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7474210690840888,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018336981744737573,
+      "loss": 1.5829,
+      "step": 2391
+    },
+    {
+      "epoch": 0.747733666770866,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001833562465631526,
+      "loss": 1.6278,
+      "step": 2392
+    },
+    {
+      "epoch": 0.748046264457643,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018334267064660668,
+      "loss": 1.6944,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7483588621444202,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018332908969855753,
+      "loss": 1.8641,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7486714598311972,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018331550371982505,
+      "loss": 1.6727,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7489840575179744,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018330191271122943,
+      "loss": 1.6077,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7492966552047515,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001832883166735912,
+      "loss": 1.4713,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7496092528915286,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018327471560773112,
+      "loss": 1.9724,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7499218505783057,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018326110951447037,
+      "loss": 1.852,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7502344482650828,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018324749839463035,
+      "loss": 1.7013,
+      "step": 2400
+    },
+    {
+      "epoch": 0.75054704595186,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018323388224903274,
+      "loss": 2.0012,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7508596436386371,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001832202610784996,
+      "loss": 1.7133,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7511722413254142,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018320663488385327,
+      "loss": 1.7841,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7514848390121913,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018319300366591637,
+      "loss": 1.8134,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7517974366989685,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018317936742551178,
+      "loss": 1.5865,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7521100343857455,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001831657261634628,
+      "loss": 1.6447,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7524226320725227,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018315207988059298,
+      "loss": 1.4747,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7527352297592997,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001831384285777261,
+      "loss": 1.7538,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7530478274460769,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018312477225568635,
+      "loss": 1.6004,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7533604251328541,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018311111091529818,
+      "loss": 1.6864,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7536730228196311,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018309744455738633,
+      "loss": 1.8215,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7539856205064083,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018308377318277587,
+      "loss": 1.672,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7542982181931853,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001830700967922921,
+      "loss": 1.7247,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7546108158799625,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018305641538676079,
+      "loss": 1.6188,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7549234135667396,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018304272896700784,
+      "loss": 1.8593,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7552360112535167,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001830290375338595,
+      "loss": 1.5332,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7555486089402939,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018301534108814234,
+      "loss": 1.5756,
+      "step": 2417
+    },
+    {
+      "epoch": 0.755861206627071,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001830016396306833,
+      "loss": 2.0714,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7561738043138481,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018298793316230948,
+      "loss": 1.64,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7564864020006252,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00018297422168384836,
+      "loss": 1.5317,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7567989996874023,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018296050519612777,
+      "loss": 1.8879,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7571115973741794,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018294678369997578,
+      "loss": 1.7005,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7574241950609566,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018293305719622072,
+      "loss": 2.244,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7577367927477336,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018291932568569134,
+      "loss": 1.5323,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7580493904345108,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018290558916921659,
+      "loss": 1.6395,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7583619881212879,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018289184764762575,
+      "loss": 1.648,
+      "step": 2426
+    },
+    {
+      "epoch": 0.758674585808065,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001828781011217485,
+      "loss": 1.5622,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7589871834948422,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018286434959241462,
+      "loss": 1.5481,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7592997811816192,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018285059306045437,
+      "loss": 1.77,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7596123788683964,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018283683152669824,
+      "loss": 1.4071,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7599249765551734,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018282306499197703,
+      "loss": 2.0644,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7602375742419506,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018280929345712186,
+      "loss": 1.7075,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7605501719287278,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001827955169229641,
+      "loss": 1.3107,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7608627696155048,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018278173539033548,
+      "loss": 1.7646,
+      "step": 2434
+    },
+    {
+      "epoch": 0.761175367302282,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018276794886006804,
+      "loss": 2.0252,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7614879649890591,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018275415733299402,
+      "loss": 1.5208,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7618005626758362,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018274036080994605,
+      "loss": 1.8906,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7621131603626133,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018272655929175708,
+      "loss": 1.8472,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7624257580493904,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001827127527792603,
+      "loss": 1.7364,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7627383557361676,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018269894127328926,
+      "loss": 1.8149,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7630509534229447,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018268512477467774,
+      "loss": 1.8335,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7633635511097218,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018267130328425985,
+      "loss": 1.7762,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7636761487964989,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018265747680287008,
+      "loss": 1.5251,
+      "step": 2443
+    },
+    {
+      "epoch": 0.763988746483276,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018264364533134304,
+      "loss": 1.5232,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7643013441700531,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018262980887051385,
+      "loss": 1.5101,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7646139418568303,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018261596742121777,
+      "loss": 1.6831,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7649265395436073,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018260212098429054,
+      "loss": 1.8748,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7652391372303845,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018258826956056793,
+      "loss": 1.7539,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7655517349171617,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018257441315088627,
+      "loss": 1.5779,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7658643326039387,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018256055175608205,
+      "loss": 1.7147,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7661769302907159,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018254668537699212,
+      "loss": 1.682,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7664895279774929,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001825328140144536,
+      "loss": 1.8002,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7668021256642701,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001825189376693039,
+      "loss": 1.3419,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7671147233510472,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001825050563423808,
+      "loss": 1.7038,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7674273210378243,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018249117003452234,
+      "loss": 1.6925,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7677399187246015,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018247727874656683,
+      "loss": 1.7601,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7680525164113785,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018246338247935285,
+      "loss": 1.6095,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7683651140981557,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001824494812337194,
+      "loss": 1.4805,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7686777117849328,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018243557501050573,
+      "loss": 1.6642,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7689903094717099,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018242166381055133,
+      "loss": 1.4541,
+      "step": 2460
+    },
+    {
+      "epoch": 0.769302907158487,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018240774763469606,
+      "loss": 1.5884,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7696155048452642,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00018239382648378006,
+      "loss": 1.6074,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7699281025320412,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018237990035864372,
+      "loss": 1.7759,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7702407002188184,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018236596926012787,
+      "loss": 1.6379,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7705532979055955,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018235203318907347,
+      "loss": 1.7159,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7708658955923726,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018233809214632184,
+      "loss": 1.6911,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7711784932791498,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018232414613271475,
+      "loss": 1.422,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7714910909659268,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018231019514909397,
+      "loss": 1.551,
+      "step": 2468
+    },
+    {
+      "epoch": 0.771803688652704,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018229623919630188,
+      "loss": 1.8121,
+      "step": 2469
+    },
+    {
+      "epoch": 0.772116286339481,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00018228227827518095,
+      "loss": 1.9086,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7724288840262582,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000182268312386574,
+      "loss": 2.5265,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7727414817130354,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001822543415313242,
+      "loss": 1.8133,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7730540793998124,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018224036571027501,
+      "loss": 1.9486,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7733666770865896,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001822263849242701,
+      "loss": 1.7464,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7736792747733667,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001822123991741536,
+      "loss": 1.743,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7739918724601438,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018219840846076977,
+      "loss": 1.4856,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7743044701469209,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018218441278496328,
+      "loss": 1.7813,
+      "step": 2477
+    },
+    {
+      "epoch": 0.774617067833698,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018217041214757903,
+      "loss": 1.7274,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7749296655204752,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018215640654946233,
+      "loss": 1.5569,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7752422632072523,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018214239599145866,
+      "loss": 1.5575,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7755548608940294,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018212838047441387,
+      "loss": 1.5972,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7758674585808065,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001821143599991741,
+      "loss": 1.668,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7761800562675836,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018210033456658576,
+      "loss": 1.646,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7764926539543607,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018208630417749561,
+      "loss": 2.3322,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7768052516411379,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018207226883275069,
+      "loss": 1.5657,
+      "step": 2485
+    },
+    {
+      "epoch": 0.777117849327915,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001820582285331983,
+      "loss": 1.4964,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7774304470146921,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018204418327968607,
+      "loss": 1.5711,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7777430447014692,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00018203013307306195,
+      "loss": 1.999,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7780556423882463,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018201607791417418,
+      "loss": 1.5581,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7783682400750235,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018200201780387126,
+      "loss": 1.5618,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7786808377618005,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018198795274300205,
+      "loss": 1.6855,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7789934354485777,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018197388273241563,
+      "loss": 1.4388,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7793060331353548,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018195980777296146,
+      "loss": 1.3961,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7796186308221319,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018194572786548924,
+      "loss": 1.3543,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7799312285089091,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018193164301084905,
+      "loss": 1.6291,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7802438261956861,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018191755320989112,
+      "loss": 1.8612,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7805564238824633,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018190345846346613,
+      "loss": 1.507,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7808690215692404,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018188935877242496,
+      "loss": 1.4034,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7811816192560175,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018187525413761887,
+      "loss": 1.3682,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7814942169427946,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018186114455989936,
+      "loss": 1.3907,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7818068146295717,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018184703004011822,
+      "loss": 1.506,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7821194123163488,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018183291057912758,
+      "loss": 1.6376,
+      "step": 2502
+    },
+    {
+      "epoch": 0.782432010003126,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018181878617777985,
+      "loss": 1.6524,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7827446076899031,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018180465683692774,
+      "loss": 1.6575,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7830572053766802,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018179052255742423,
+      "loss": 1.6608,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7833698030634574,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00018177638334012267,
+      "loss": 1.7274,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7836824007502344,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018176223918587664,
+      "loss": 1.7459,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7839949984370116,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018174809009554005,
+      "loss": 1.366,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7843075961237886,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018173393606996707,
+      "loss": 1.7907,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7846201938105658,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001817197771100122,
+      "loss": 1.7705,
+      "step": 2510
+    },
+    {
+      "epoch": 0.784932791497343,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00018170561321653026,
+      "loss": 1.4995,
+      "step": 2511
+    },
+    {
+      "epoch": 0.78524538918412,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018169144439037632,
+      "loss": 1.6226,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7855579868708972,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018167727063240582,
+      "loss": 1.619,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7858705845576742,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018166309194347438,
+      "loss": 1.9021,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7861831822444514,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000181648908324438,
+      "loss": 1.9489,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7864957799312285,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018163471977615303,
+      "loss": 1.5399,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7868083776180056,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000181620526299476,
+      "loss": 1.5515,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7871209753047828,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018160632789526374,
+      "loss": 1.4493,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7874335729915599,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018159212456437347,
+      "loss": 1.6494,
+      "step": 2519
+    },
+    {
+      "epoch": 0.787746170678337,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001815779163076627,
+      "loss": 1.7547,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7880587683651141,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018156370312598914,
+      "loss": 1.7275,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7883713660518912,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0001815494850202109,
+      "loss": 1.3418,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7886839637386683,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00018153526199118634,
+      "loss": 1.5102,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7889965614254455,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001815210340397741,
+      "loss": 2.0452,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7893091591122225,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00018150680116683313,
+      "loss": 1.5017,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7896217567989997,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018149256337322275,
+      "loss": 2.0215,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7899343544857768,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018147832065980245,
+      "loss": 1.7694,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7902469521725539,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018146407302743208,
+      "loss": 1.6186,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7905595498593311,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018144982047697185,
+      "loss": 1.7227,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7908721475461081,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018143556300928215,
+      "loss": 1.6313,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7911847452328853,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018142130062522377,
+      "loss": 1.4294,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7914973429196624,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018140703332565768,
+      "loss": 1.5747,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7918099406064395,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018139276111144525,
+      "loss": 1.6087,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7921225382932167,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001813784839834481,
+      "loss": 1.6986,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7924351359799937,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018136420194252818,
+      "loss": 1.5952,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7927477336667709,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00018134991498954773,
+      "loss": 1.7808,
+      "step": 2536
+    },
+    {
+      "epoch": 0.793060331353548,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001813356231253692,
+      "loss": 1.518,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7933729290403251,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001813213263508555,
+      "loss": 1.82,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7936855267271022,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001813070246668697,
+      "loss": 1.5595,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7939981244138793,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018129271807427517,
+      "loss": 1.8371,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7943107221006565,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001812784065739357,
+      "loss": 1.5297,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7946233197874336,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001812640901667152,
+      "loss": 1.6262,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7949359174742107,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018124976885347806,
+      "loss": 1.7128,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7952485151609878,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018123544263508884,
+      "loss": 1.9219,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7955611128477649,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018122111151241241,
+      "loss": 1.5844,
+      "step": 2545
+    },
+    {
+      "epoch": 0.795873710534542,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000181206775486314,
+      "loss": 1.806,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7961863082213192,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018119243455765903,
+      "loss": 1.648,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7964989059080962,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018117808872731336,
+      "loss": 1.5256,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7968115035948734,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000181163737996143,
+      "loss": 1.491,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7971241012816506,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00018114938236501438,
+      "loss": 1.8205,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7974366989684276,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001811350218347941,
+      "loss": 1.6017,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7977492966552048,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001811206564063492,
+      "loss": 1.4423,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7980618943419818,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018110628608054686,
+      "loss": 1.8525,
+      "step": 2553
+    },
+    {
+      "epoch": 0.798374492028759,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001810919108582547,
+      "loss": 1.7098,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7986870897155361,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018107753074034054,
+      "loss": 1.7347,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7989996874023132,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018106314572767252,
+      "loss": 1.6353,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7993122850890904,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018104875582111913,
+      "loss": 1.7014,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7996248827758674,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00018103436102154903,
+      "loss": 1.5313,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7999374804626446,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001810199613298313,
+      "loss": 1.671,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8002500781494217,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00018100555674683527,
+      "loss": 1.5859,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8005626758361988,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018099114727343057,
+      "loss": 1.4992,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8008752735229759,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00018097673291048706,
+      "loss": 1.6654,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8011878712097531,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000180962313658875,
+      "loss": 1.6192,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8015004688965301,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001809478895194649,
+      "loss": 1.7311,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8018130665833073,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018093346049312758,
+      "loss": 1.5685,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8021256642700844,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001809190265807341,
+      "loss": 1.9562,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8024382619568615,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018090458778315588,
+      "loss": 1.662,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8027508596436387,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018089014410126457,
+      "loss": 1.611,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8030634573304157,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001808756955359322,
+      "loss": 1.7113,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8033760550171929,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018086124208803103,
+      "loss": 1.3589,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8036886527039699,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018084678375843364,
+      "loss": 1.819,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8040012503907471,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018083232054801288,
+      "loss": 1.6764,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8043138480775243,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001808178524576419,
+      "loss": 1.5922,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8046264457643013,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001808033794881942,
+      "loss": 1.5336,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8049390434510785,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0001807889016405435,
+      "loss": 1.443,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8052516411378556,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001807744189155639,
+      "loss": 1.7123,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8055642388246327,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018075993131412966,
+      "loss": 1.9127,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8058768365114098,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018074543883711547,
+      "loss": 1.7716,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8061894341981869,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00018073094148539625,
+      "loss": 1.7905,
+      "step": 2579
+    },
+    {
+      "epoch": 0.806502031884964,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00018071643925984717,
+      "loss": 1.5217,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8068146295717412,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018070193216134384,
+      "loss": 1.6451,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8071272272585183,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018068742019076203,
+      "loss": 1.7439,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8074398249452954,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001806729033489778,
+      "loss": 2.0439,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8077524226320725,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001806583816368676,
+      "loss": 1.7726,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8080650203188496,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018064385505530813,
+      "loss": 1.8142,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8083776180056268,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018062932360517637,
+      "loss": 1.8507,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8086902156924038,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001806147872873496,
+      "loss": 1.8861,
+      "step": 2587
+    },
+    {
+      "epoch": 0.809002813379181,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018060024610270538,
+      "loss": 2.04,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8093154110659581,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001805857000521216,
+      "loss": 1.5433,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8096280087527352,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00018057114913647642,
+      "loss": 1.5803,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8099406064395124,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001805565933566483,
+      "loss": 1.7928,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8102532041262894,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00018054203271351599,
+      "loss": 1.8568,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8105658018130666,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00018052746720795848,
+      "loss": 1.5727,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8108783994998437,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00018051289684085518,
+      "loss": 1.543,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8111909971866208,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018049832161308574,
+      "loss": 1.5196,
+      "step": 2595
+    },
+    {
+      "epoch": 0.811503594873398,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018048374152553,
+      "loss": 1.592,
+      "step": 2596
+    },
+    {
+      "epoch": 0.811816192560175,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018046915657906826,
+      "loss": 1.6238,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8121287902469522,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018045456677458094,
+      "loss": 1.6494,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8124413879337293,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018043997211294896,
+      "loss": 1.7159,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8127539856205064,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018042537259505332,
+      "loss": 1.7333,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8130665833072835,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018041076822177546,
+      "loss": 1.7428,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8133791809940606,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00018039615899399704,
+      "loss": 1.5266,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8136917786808378,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00018038154491260006,
+      "loss": 1.4482,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8140043763676149,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001803669259784668,
+      "loss": 1.8164,
+      "step": 2604
+    },
+    {
+      "epoch": 0.814316974054392,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018035230219247978,
+      "loss": 1.7801,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8146295717411691,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001803376735555219,
+      "loss": 1.5818,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8149421694279463,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001803230400684763,
+      "loss": 2.0025,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8152547671147233,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001803084017322264,
+      "loss": 1.6328,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8155673648015005,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00018029375854765597,
+      "loss": 2.289,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8158799624882775,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018027911051564897,
+      "loss": 1.4681,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8161925601750547,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001802644576370898,
+      "loss": 1.7437,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8165051578618319,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018024979991286303,
+      "loss": 2.0136,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8168177555486089,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001802351373438536,
+      "loss": 1.6401,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8171303532353861,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00018022046993094665,
+      "loss": 1.5986,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8174429509221631,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00018020579767502774,
+      "loss": 1.7392,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8177555486089403,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001801911205769826,
+      "loss": 1.6622,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8180681462957174,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001801764386376973,
+      "loss": 1.6786,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8183807439824945,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001801617518580583,
+      "loss": 1.6723,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8186933416692717,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001801470602389521,
+      "loss": 1.6344,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8190059393560488,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001801323637812658,
+      "loss": 1.8773,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8193185370428259,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00018011766248588655,
+      "loss": 1.7633,
+      "step": 2621
+    },
+    {
+      "epoch": 0.819631134729603,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00018010295635370192,
+      "loss": 1.7818,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8199437324163801,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018008824538559977,
+      "loss": 1.5338,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8202563301031572,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00018007352958246818,
+      "loss": 2.1521,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8205689277899344,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00018005880894519555,
+      "loss": 1.6819,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8208815254767114,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00018004408347467062,
+      "loss": 1.7966,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8211941231634886,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018002935317178235,
+      "loss": 1.5681,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8215067208502657,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00018001461803742008,
+      "loss": 1.8119,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8218193185370428,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017999987807247334,
+      "loss": 2.2241,
+      "step": 2629
+    },
+    {
+      "epoch": 0.82213191622382,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017998513327783199,
+      "loss": 1.5033,
+      "step": 2630
+    },
+    {
+      "epoch": 0.822444513910597,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017997038365438628,
+      "loss": 1.481,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8227571115973742,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017995562920302652,
+      "loss": 1.7684,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8230697092841514,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001799408699246436,
+      "loss": 1.6599,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8233823069709284,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017992610582012847,
+      "loss": 1.3327,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8236949046577056,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001799113368903725,
+      "loss": 1.7121,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8240075023444826,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00017989656313626727,
+      "loss": 1.766,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8243201000312598,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001798817845587047,
+      "loss": 1.85,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8246326977180369,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000179867001158577,
+      "loss": 1.8962,
+      "step": 2638
+    },
+    {
+      "epoch": 0.824945295404814,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001798522129367767,
+      "loss": 1.4497,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8252578930915911,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017983741989419655,
+      "loss": 1.6794,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8255704907783682,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001798226220317296,
+      "loss": 1.718,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8258830884651454,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017980781935026925,
+      "loss": 1.7489,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8261956861519225,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001797930118507091,
+      "loss": 1.7344,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8265082838386996,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001797781995339432,
+      "loss": 1.7674,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8268208815254767,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001797633824008657,
+      "loss": 2.0352,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8271334792122538,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00017974856045237117,
+      "loss": 1.6354,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8274460768990309,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017973373368935445,
+      "loss": 1.737,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8277586745858081,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017971890211271059,
+      "loss": 1.7081,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8280712722725851,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000179704065723335,
+      "loss": 1.3865,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8283838699593623,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017968922452212343,
+      "loss": 1.5347,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8286964676461395,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017967437850997185,
+      "loss": 1.7372,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8290090653329165,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017965952768777649,
+      "loss": 1.5994,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8293216630196937,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001796446720564339,
+      "loss": 1.8905,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8296342607064707,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00017962981161684098,
+      "loss": 2.5074,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8299468583932479,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017961494636989486,
+      "loss": 1.9347,
+      "step": 2655
+    },
+    {
+      "epoch": 0.830259456080025,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017960007631649298,
+      "loss": 1.8819,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8305720537668021,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017958520145753307,
+      "loss": 1.6299,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8308846514535793,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017957032179391312,
+      "loss": 1.7028,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8311972491403563,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017955543732653143,
+      "loss": 1.8788,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8315098468271335,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001795405480562866,
+      "loss": 1.7432,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8318224445139106,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017952565398407757,
+      "loss": 1.583,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8321350422006877,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017951075511080347,
+      "loss": 1.7078,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8324476398874648,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001794958514373637,
+      "loss": 1.4488,
+      "step": 2663
+    },
+    {
+      "epoch": 0.832760237574242,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017948094296465814,
+      "loss": 1.6082,
+      "step": 2664
+    },
+    {
+      "epoch": 0.833072835261019,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017946602969358673,
+      "loss": 1.6088,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8333854329477962,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017945111162504987,
+      "loss": 1.7525,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8336980306345733,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017943618875994815,
+      "loss": 1.8168,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8340106283213504,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017942126109918248,
+      "loss": 1.7631,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8343232260081276,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017940632864365408,
+      "loss": 1.665,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8346358236949046,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00017939139139426443,
+      "loss": 1.7743,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8349484213816818,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001793764493519153,
+      "loss": 1.6251,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8352610190684588,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017936150251750876,
+      "loss": 1.5676,
+      "step": 2672
+    },
+    {
+      "epoch": 0.835573616755236,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001793465508919472,
+      "loss": 1.9198,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8358862144420132,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017933159447613325,
+      "loss": 1.8999,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8361988121287902,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017931663327096985,
+      "loss": 1.5773,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8365114098155674,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017930166727736022,
+      "loss": 1.5615,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8368240075023445,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001792866964962079,
+      "loss": 1.7466,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8371366051891216,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017927172092841665,
+      "loss": 1.5719,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8374492028758987,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017925674057489062,
+      "loss": 1.8351,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8377618005626758,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017924175543653412,
+      "loss": 1.3423,
+      "step": 2680
+    },
+    {
+      "epoch": 0.838074398249453,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001792267655142519,
+      "loss": 1.8691,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8383869959362301,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017921177080894887,
+      "loss": 1.5727,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8386995936230072,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001791967713215303,
+      "loss": 1.5138,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8390121913097843,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017918176705290174,
+      "loss": 1.7783,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8393247889965614,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017916675800396897,
+      "loss": 1.8948,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8396373866833385,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017915174417563816,
+      "loss": 1.6654,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8399499843701157,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017913672556881566,
+      "loss": 1.8393,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8402625820568927,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017912170218440822,
+      "loss": 1.5724,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8405751797436699,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001791066740233228,
+      "loss": 1.5801,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8408877774304471,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017909164108646667,
+      "loss": 1.6645,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8412003751172241,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017907660337474735,
+      "loss": 1.6794,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8415129728040013,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001790615608890727,
+      "loss": 1.6382,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8418255704907783,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017904651363035093,
+      "loss": 1.6977,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8421381681775555,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017903146159949036,
+      "loss": 1.4432,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8424507658643327,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00017901640479739975,
+      "loss": 1.7628,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8427633635511097,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001790013432249881,
+      "loss": 1.6406,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8430759612378869,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017898627688316468,
+      "loss": 2.2605,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8433885589246639,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017897120577283908,
+      "loss": 1.6559,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8437011566114411,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017895612989492113,
+      "loss": 1.7878,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8440137542982182,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000178941049250321,
+      "loss": 1.7082,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8443263519849953,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00017892596383994915,
+      "loss": 1.6265,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8446389496717724,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017891087366471632,
+      "loss": 1.6036,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8449515473585495,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017889577872553343,
+      "loss": 1.4701,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8452641450453267,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017888067902331186,
+      "loss": 1.7345,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8455767427321038,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001788655745589632,
+      "loss": 1.7042,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8458893404188809,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001788504653333993,
+      "loss": 1.9033,
+      "step": 2706
+    },
+    {
+      "epoch": 0.846201938105658,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0001788353513475323,
+      "loss": 1.6525,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8465145357924352,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001788202326022747,
+      "loss": 1.6119,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8468271334792122,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001788051090985392,
+      "loss": 1.7473,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8471397311659894,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017878998083723885,
+      "loss": 1.8992,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8474523288527664,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017877484781928698,
+      "loss": 1.6285,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8477649265395436,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017875971004559712,
+      "loss": 1.671,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8480775242263208,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001787445675170832,
+      "loss": 1.639,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8483901219130978,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017872942023465944,
+      "loss": 2.2887,
+      "step": 2714
+    },
+    {
+      "epoch": 0.848702719599875,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017871426819924025,
+      "loss": 1.6424,
+      "step": 2715
+    },
+    {
+      "epoch": 0.849015317286652,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017869911141174034,
+      "loss": 1.6615,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8493279149734292,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017868394987307482,
+      "loss": 1.8865,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8496405126602063,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017866878358415895,
+      "loss": 1.4584,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8499531103469834,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001786536125459084,
+      "loss": 1.7852,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8502657080337606,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000178638436759239,
+      "loss": 1.5773,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8505783057205377,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017862325622506698,
+      "loss": 1.5571,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8508909034073148,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017860807094430877,
+      "loss": 1.6325,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8512035010940919,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001785928809178812,
+      "loss": 2.1872,
+      "step": 2723
+    },
+    {
+      "epoch": 0.851516098780869,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001785776861467012,
+      "loss": 1.7218,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8518286964676461,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017856248663168618,
+      "loss": 1.8967,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8521412941544233,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017854728237375373,
+      "loss": 1.412,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8524538918412004,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017853207337382174,
+      "loss": 1.5824,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8527664895279775,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001785168596328084,
+      "loss": 1.6068,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8530790872147546,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001785016411516322,
+      "loss": 1.5164,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8533916849015317,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017848641793121188,
+      "loss": 1.8491,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8537042825883089,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001784711899724665,
+      "loss": 1.6247,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8540168802750859,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001784559572763154,
+      "loss": 1.4966,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8543294779618631,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00017844071984367816,
+      "loss": 1.5311,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8546420756486403,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001784254776754747,
+      "loss": 1.521,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8549546733354173,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017841023077262523,
+      "loss": 1.7637,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8552672710221945,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001783949791360502,
+      "loss": 1.3663,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8555798687089715,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001783797227666704,
+      "loss": 1.6854,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8558924663957487,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017836446166540683,
+      "loss": 1.7461,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8562050640825258,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017834919583318087,
+      "loss": 1.5579,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8565176617693029,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017833392527091412,
+      "loss": 1.8503,
+      "step": 2740
+    },
+    {
+      "epoch": 0.85683025945608,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017831864997952846,
+      "loss": 1.7036,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017830336995994608,
+      "loss": 1.546,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8574554548296343,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017828808521308949,
+      "loss": 1.7367,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8577680525164114,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017827279573988145,
+      "loss": 1.6342,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8580806502031885,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017825750154124497,
+      "loss": 1.4992,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8583932478899656,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017824220261810337,
+      "loss": 1.6274,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8587058455767427,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017822689897138035,
+      "loss": 1.4625,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8590184432635198,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017821159060199974,
+      "loss": 1.4388,
+      "step": 2748
+    },
+    {
+      "epoch": 0.859331040950297,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017819627751088573,
+      "loss": 1.4505,
+      "step": 2749
+    },
+    {
+      "epoch": 0.859643638637074,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0001781809596989628,
+      "loss": 1.4593,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8599562363238512,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001781656371671557,
+      "loss": 1.5498,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8602688340106284,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00017815030991638947,
+      "loss": 2.1876,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8605814316974054,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017813497794758946,
+      "loss": 1.4955,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8608940293841826,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017811964126168123,
+      "loss": 1.6525,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8612066270709596,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017810429985959077,
+      "loss": 1.7273,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8615192247577368,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017808895374224414,
+      "loss": 1.6337,
+      "step": 2756
+    },
+    {
+      "epoch": 0.861831822444514,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001780736029105679,
+      "loss": 1.572,
+      "step": 2757
+    },
+    {
+      "epoch": 0.862144420131291,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017805824736548872,
+      "loss": 1.7677,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8624570178180682,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017804288710793374,
+      "loss": 1.4813,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8627696155048452,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017802752213883017,
+      "loss": 1.863,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8630822131916224,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017801215245910569,
+      "loss": 1.7106,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8633948108783995,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017799677806968811,
+      "loss": 1.5748,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8637074085651766,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017798139897150564,
+      "loss": 1.7248,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8640200062519537,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017796601516548676,
+      "loss": 1.7132,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8643326039387309,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001779506266525602,
+      "loss": 1.742,
+      "step": 2765
+    },
+    {
+      "epoch": 0.864645201625508,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.000177935233433655,
+      "loss": 1.8706,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8649577993122851,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001779198355097004,
+      "loss": 1.5686,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8652703969990622,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017790443288162605,
+      "loss": 1.7863,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8655829946858393,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017788902555036182,
+      "loss": 1.6466,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8658955923726165,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017787361351683786,
+      "loss": 1.7133,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8662081900593935,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00017785819678198462,
+      "loss": 1.7669,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8665207877461707,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001778427753467328,
+      "loss": 1.7054,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8668333854329477,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017782734921201348,
+      "loss": 1.5878,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8671459831197249,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017781191837875788,
+      "loss": 1.5847,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8674585808065021,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001777964828478976,
+      "loss": 1.556,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8677711784932791,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017778104262036455,
+      "loss": 1.481,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8680837761800563,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001777655976970908,
+      "loss": 1.5842,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8683963738668334,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017775014807900884,
+      "loss": 1.6188,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8687089715536105,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017773469376705138,
+      "loss": 1.7405,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8690215692403876,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017771923476215138,
+      "loss": 2.009,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8693341669271647,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017770377106524215,
+      "loss": 1.5022,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8696467646139419,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0001776883026772572,
+      "loss": 2.3243,
+      "step": 2782
+    },
+    {
+      "epoch": 0.869959362300719,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017767282959913047,
+      "loss": 1.5778,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8702719599874961,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017765735183179602,
+      "loss": 1.648,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8705845576742732,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017764186937618828,
+      "loss": 1.9461,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8708971553610503,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017762638223324192,
+      "loss": 1.6331,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8712097530478274,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017761089040389198,
+      "loss": 1.5506,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8715223507346046,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017759539388907366,
+      "loss": 1.4817,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8718349484213817,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017757989268972257,
+      "loss": 1.4606,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8721475461081588,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017756438680677445,
+      "loss": 1.4484,
+      "step": 2790
+    },
+    {
+      "epoch": 0.872460143794936,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017754887624116548,
+      "loss": 1.5865,
+      "step": 2791
+    },
+    {
+      "epoch": 0.872772741481713,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017753336099383203,
+      "loss": 1.514,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8730853391684902,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017751784106571079,
+      "loss": 1.3963,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8733979368552672,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017750231645773869,
+      "loss": 1.8982,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8737105345420444,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017748678717085297,
+      "loss": 1.7107,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8740231322288216,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017747125320599118,
+      "loss": 1.5219,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8743357299155986,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001774557145640911,
+      "loss": 1.6148,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8746483276023758,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017744017124609083,
+      "loss": 1.4968,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8749609252891528,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017742462325292873,
+      "loss": 1.6438,
+      "step": 2799
+    },
+    {
+      "epoch": 0.87527352297593,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001774090705855435,
+      "loss": 1.8157,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8755861206627071,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.000177393513244874,
+      "loss": 1.8969,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8758987183494842,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001773779512318595,
+      "loss": 1.7561,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8762113160362613,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017736238454743946,
+      "loss": 1.8387,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8765239137230384,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001773468131925537,
+      "loss": 1.8426,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8768365114098156,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017733123716814225,
+      "loss": 1.5613,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8771491090965927,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001773156564751455,
+      "loss": 1.9907,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8774617067833698,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017730007111450402,
+      "loss": 1.3814,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8777743044701469,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017728448108715874,
+      "loss": 1.459,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8780869021569241,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017726888639405086,
+      "loss": 1.6541,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8783994998437011,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017725328703612183,
+      "loss": 1.6136,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8787120975304783,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017723768301431344,
+      "loss": 1.9023,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8790246952172553,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00017722207432956767,
+      "loss": 2.4062,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8793372929040325,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017720646098282687,
+      "loss": 1.6481,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8796498905908097,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017719084297503367,
+      "loss": 1.7955,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8799624882775867,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001771752203071309,
+      "loss": 1.7442,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8802750859643639,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001771595929800617,
+      "loss": 1.9734,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8805876836511409,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001771439609947696,
+      "loss": 1.651,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8809002813379181,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017712832435219823,
+      "loss": 1.6914,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8812128790246953,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017711268305329166,
+      "loss": 1.9028,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8815254767114723,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017709703709899413,
+      "loss": 1.7345,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8818380743982495,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017708138649025023,
+      "loss": 1.8512,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8821506720850266,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001770657312280048,
+      "loss": 1.6781,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8824632697718037,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017705007131320298,
+      "loss": 1.5084,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8827758674585808,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017703440674679015,
+      "loss": 1.5801,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8830884651453579,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00017701873752971206,
+      "loss": 1.7738,
+      "step": 2825
+    },
+    {
+      "epoch": 0.883401062832135,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017700306366291458,
+      "loss": 1.7093,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8837136605189122,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017698738514734406,
+      "loss": 1.7994,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8840262582056893,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00017697170198394696,
+      "loss": 1.7524,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8843388558924664,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001769560141736702,
+      "loss": 1.4667,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8846514535792435,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017694032171746072,
+      "loss": 1.4843,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8849640512660206,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000176924624616266,
+      "loss": 1.4988,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8852766489527978,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017690892287103367,
+      "loss": 1.5816,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8855892466395748,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017689321648271166,
+      "loss": 1.7245,
+      "step": 2833
+    },
+    {
+      "epoch": 0.885901844326352,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00017687750545224815,
+      "loss": 1.7804,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8862144420131292,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001768617897805917,
+      "loss": 1.5097,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8865270396999062,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017684606946869106,
+      "loss": 1.5496,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8868396373866834,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017683034451749526,
+      "loss": 1.829,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8871522350734604,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001768146149279537,
+      "loss": 1.4844,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8874648327602376,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017679888070101592,
+      "loss": 1.7066,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8877774304470147,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017678314183763183,
+      "loss": 1.5307,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8880900281337918,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017676739833875164,
+      "loss": 1.4304,
+      "step": 2841
+    },
+    {
+      "epoch": 0.888402625820569,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00017675165020532578,
+      "loss": 1.6068,
+      "step": 2842
+    },
+    {
+      "epoch": 0.888715223507346,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000176735897438305,
+      "loss": 1.4709,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8890278211941232,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017672014003864033,
+      "loss": 1.6562,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8893404188809003,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000176704378007283,
+      "loss": 1.8352,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8896530165676774,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0001766886113451846,
+      "loss": 1.8639,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8899656142544545,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017667284005329708,
+      "loss": 1.6163,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8902782119412317,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017665706413257245,
+      "loss": 1.7933,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8905908096280087,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001766412835839632,
+      "loss": 1.6013,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8909034073147859,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000176625498408422,
+      "loss": 1.6694,
+      "step": 2850
+    },
+    {
+      "epoch": 0.891216005001563,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001766097086069018,
+      "loss": 1.6816,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8915286026883401,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017659391418035588,
+      "loss": 1.7289,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8918412003751173,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001765781151297377,
+      "loss": 1.4146,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8921537980618943,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001765623114560012,
+      "loss": 1.6338,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8924663957486715,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017654650316010036,
+      "loss": 1.623,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8927789934354485,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017653069024298957,
+      "loss": 1.6547,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8930915911222257,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0001765148727056235,
+      "loss": 1.7697,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8934041888090029,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017649905054895705,
+      "loss": 1.7488,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8937167864957799,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017648322377394546,
+      "loss": 2.1237,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8940293841825571,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017646739238154417,
+      "loss": 1.6839,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8943419818693341,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017645155637270897,
+      "loss": 1.6423,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8946545795561113,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017643571574839587,
+      "loss": 1.7184,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8949671772428884,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017641987050956122,
+      "loss": 1.8355,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8952797749296655,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0001764040206571616,
+      "loss": 1.6686,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8955923726164426,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017638816619215388,
+      "loss": 1.7545,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8959049703032198,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017637230711549525,
+      "loss": 1.7738,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8962175679899969,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001763564434281431,
+      "loss": 1.7099,
+      "step": 2867
+    },
+    {
+      "epoch": 0.896530165676774,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017634057513105515,
+      "loss": 1.6731,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8968427633635511,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001763247022251894,
+      "loss": 1.4654,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8971553610503282,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017630882471150413,
+      "loss": 1.7359,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8974679587371054,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017629294259095785,
+      "loss": 1.5702,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8977805564238824,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017627705586450944,
+      "loss": 2.429,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8980931541106596,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017626116453311794,
+      "loss": 1.8714,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8984057517974366,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017624526859774274,
+      "loss": 1.592,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8987183494842138,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017622936805934355,
+      "loss": 1.9351,
+      "step": 2875
+    },
+    {
+      "epoch": 0.899030947170991,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017621346291888025,
+      "loss": 1.5676,
+      "step": 2876
+    },
+    {
+      "epoch": 0.899343544857768,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001761975531773131,
+      "loss": 2.0676,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8996561425445452,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017618163883560255,
+      "loss": 1.8676,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8999687402313223,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017616571989470937,
+      "loss": 1.6823,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9002813379180994,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017614979635559462,
+      "loss": 1.6829,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9005939356048765,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017613386821921964,
+      "loss": 1.3811,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9009065332916536,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017611793548654602,
+      "loss": 1.3734,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9012191309784308,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00017610199815853563,
+      "loss": 1.8464,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9015317286652079,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017608605623615063,
+      "loss": 1.4275,
+      "step": 2884
+    },
+    {
+      "epoch": 0.901844326351985,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017607010972035348,
+      "loss": 1.5875,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9021569240387621,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017605415861210685,
+      "loss": 1.8575,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9024695217255392,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017603820291237375,
+      "loss": 1.8156,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9027821194123163,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017602224262211743,
+      "loss": 1.4908,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9030947170990935,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017600627774230144,
+      "loss": 1.7584,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9034073147858706,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017599030827388965,
+      "loss": 1.7706,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9037199124726477,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001759743342178461,
+      "loss": 1.6771,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9040325101594249,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017595835557513516,
+      "loss": 1.838,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9043451078462019,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017594237234672152,
+      "loss": 1.7833,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9046577055329791,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017592638453357005,
+      "loss": 1.8564,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9049703032197561,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000175910392136646,
+      "loss": 1.4054,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9052829009065333,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017589439515691487,
+      "loss": 1.7344,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9055954985933105,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001758783935953424,
+      "loss": 1.6391,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9059080962800875,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017586238745289457,
+      "loss": 1.6244,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9062206939668647,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017584637673053778,
+      "loss": 1.6056,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9065332916536417,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017583036142923856,
+      "loss": 1.7858,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9068458893404189,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001758143415499638,
+      "loss": 1.6028,
+      "step": 2901
+    },
+    {
+      "epoch": 0.907158487027196,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001757983170936806,
+      "loss": 1.6918,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9074710847139731,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017578228806135643,
+      "loss": 1.9901,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9077836824007502,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017576625445395893,
+      "loss": 1.5383,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9080962800875274,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017575021627245612,
+      "loss": 1.5068,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9084088777743045,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017573417351781625,
+      "loss": 1.8062,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9087214754610816,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00017571812619100778,
+      "loss": 1.4791,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9090340731478587,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017570207429299956,
+      "loss": 1.7496,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9093466708346358,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017568601782476064,
+      "loss": 1.5202,
+      "step": 2909
+    },
+    {
+      "epoch": 0.909659268521413,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017566995678726038,
+      "loss": 1.6579,
+      "step": 2910
+    },
+    {
+      "epoch": 0.90997186620819,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001756538911814684,
+      "loss": 1.606,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9102844638949672,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001756378210083546,
+      "loss": 1.6417,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9105970615817442,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017562174626888918,
+      "loss": 1.6654,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9109096592685214,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017560566696404254,
+      "loss": 1.676,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9112222569552986,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017558958309478543,
+      "loss": 1.5845,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9115348546420756,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001755734946620889,
+      "loss": 1.5907,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9118474523288528,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017555740166692418,
+      "loss": 1.8526,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9121600500156298,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017554130411026283,
+      "loss": 1.4743,
+      "step": 2918
+    },
+    {
+      "epoch": 0.912472647702407,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001755252019930767,
+      "loss": 1.4929,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9127852453891842,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001755090953163379,
+      "loss": 1.4583,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9130978430759612,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017549298408101876,
+      "loss": 1.7967,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9134104407627384,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017547686828809196,
+      "loss": 1.9172,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9137230384495155,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017546074793853048,
+      "loss": 1.5975,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9140356361362926,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017544462303330748,
+      "loss": 1.8838,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9143482338230697,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017542849357339644,
+      "loss": 1.8619,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9146608315098468,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017541235955977112,
+      "loss": 1.6366,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9149734291966239,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017539622099340554,
+      "loss": 1.5817,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9152860268834011,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000175380077875274,
+      "loss": 1.5323,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9155986245701782,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017536393020635118,
+      "loss": 1.762,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9159112222569553,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001753477779876118,
+      "loss": 1.5217,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9162238199437324,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017533162122003107,
+      "loss": 1.6377,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9165364176305095,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017531545990458436,
+      "loss": 1.5614,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9168490153172867,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017529929404224733,
+      "loss": 1.9785,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9171616130040637,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017528312363399598,
+      "loss": 1.6278,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9174742106908409,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017526694868080656,
+      "loss": 1.62,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9177868083776181,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001752507691836555,
+      "loss": 1.66,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9180994060643951,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017523458514351963,
+      "loss": 1.711,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9184120037511723,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017521839656137598,
+      "loss": 1.606,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9187246014379493,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017520220343820184,
+      "loss": 1.8548,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9190371991247265,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017518600577497487,
+      "loss": 1.6217,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9193497968115036,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00017516980357267295,
+      "loss": 2.4887,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9196623944982807,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017515359683227416,
+      "loss": 1.7841,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9199749921850578,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00017513738555475697,
+      "loss": 1.7065,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9202875898718349,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001751211697411001,
+      "loss": 1.7469,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9206001875586121,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00017510494939228246,
+      "loss": 1.5839,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9209127852453892,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001750887245092833,
+      "loss": 1.7413,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9212253829321663,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017507249509308217,
+      "loss": 1.433,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9215379806189434,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017505626114465886,
+      "loss": 1.5907,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9218505783057206,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001750400226649934,
+      "loss": 1.6737,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9221631759924976,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017502377965506613,
+      "loss": 1.5084,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9224757736792748,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017500753211585772,
+      "loss": 1.4999,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9227883713660519,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.000174991280048349,
+      "loss": 1.6843,
+      "step": 2952
+    },
+    {
+      "epoch": 0.923100969052829,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017497502345352112,
+      "loss": 1.6222,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9234135667396062,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017495876233235554,
+      "loss": 1.5935,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9237261644263832,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000174942496685834,
+      "loss": 1.9363,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9240387621131604,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017492622651493837,
+      "loss": 1.8212,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9243513597999374,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000174909951820651,
+      "loss": 1.8014,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9246639574867146,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017489367260395438,
+      "loss": 1.7982,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9249765551734918,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001748773888658313,
+      "loss": 1.6039,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9252891528602688,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017486110060726485,
+      "loss": 1.8941,
+      "step": 2960
+    },
+    {
+      "epoch": 0.925601750547046,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017484480782923835,
+      "loss": 2.0574,
+      "step": 2961
+    },
+    {
+      "epoch": 0.925914348233823,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017482851053273542,
+      "loss": 1.404,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9262269459206002,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00017481220871873996,
+      "loss": 1.6843,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9265395436073773,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017479590238823613,
+      "loss": 1.61,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9268521412941544,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017477959154220835,
+      "loss": 2.4723,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9271647389809315,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001747632761816413,
+      "loss": 1.6597,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9274773366677087,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017474695630752008,
+      "loss": 1.5784,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9277899343544858,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00017473063192082982,
+      "loss": 1.8403,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9281025320412629,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017471430302255604,
+      "loss": 1.8024,
+      "step": 2969
+    },
+    {
+      "epoch": 0.92841512972804,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017469796961368462,
+      "loss": 1.714,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9287277274148171,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017468163169520156,
+      "loss": 1.4359,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9290403251015943,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017466528926809324,
+      "loss": 1.6177,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9293529227883713,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017464894233334627,
+      "loss": 1.9172,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9296655204751485,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017463259089194752,
+      "loss": 2.023,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9299781181619255,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017461623494488416,
+      "loss": 1.3345,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9302907158487027,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001745998744931436,
+      "loss": 1.6451,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9306033135354799,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017458350953771355,
+      "loss": 1.4398,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9309159112222569,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000174567140079582,
+      "loss": 1.4698,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9312285089090341,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017455076611973716,
+      "loss": 1.586,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9315411065958112,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017453438765916758,
+      "loss": 1.4608,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9318537042825883,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017451800469886207,
+      "loss": 1.7327,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9321663019693655,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001745016172398096,
+      "loss": 1.7701,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9324788996561425,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001744852252829996,
+      "loss": 1.6054,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9327914973429197,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017446882882942162,
+      "loss": 1.7484,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9331040950296968,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017445242788006552,
+      "loss": 1.6647,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9334166927164739,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001744360224359215,
+      "loss": 1.6536,
+      "step": 2986
+    },
+    {
+      "epoch": 0.933729290403251,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017441961249797995,
+      "loss": 1.9033,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9340418880900281,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017440319806723157,
+      "loss": 1.5145,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9343544857768052,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001743867791446673,
+      "loss": 1.6766,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9346670834635824,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017437035573127836,
+      "loss": 1.5665,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9349796811503595,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017435392782805628,
+      "loss": 1.7932,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9352922788371366,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017433749543599287,
+      "loss": 1.595,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9356048765239138,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00017432105855608008,
+      "loss": 1.7333,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9359174742106908,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001743046171893103,
+      "loss": 1.6385,
+      "step": 2994
+    },
+    {
+      "epoch": 0.936230071897468,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001742881713366761,
+      "loss": 1.7989,
+      "step": 2995
+    },
+    {
+      "epoch": 0.936542669584245,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017427172099917032,
+      "loss": 1.5065,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9368552672710222,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001742552661777861,
+      "loss": 1.6564,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9371678649577994,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017423880687351685,
+      "loss": 1.5779,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9374804626445764,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001742223430873562,
+      "loss": 1.7974,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9377930603313536,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001742058748202981,
+      "loss": 1.4744,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9381056580181306,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001741894020733368,
+      "loss": 1.6008,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9384182557049078,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017417292484746676,
+      "loss": 1.5435,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9387308533916849,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017415644314368274,
+      "loss": 1.6641,
+      "step": 3003
+    },
+    {
+      "epoch": 0.939043451078462,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017413995696297972,
+      "loss": 1.661,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9393560487652391,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017412346630635303,
+      "loss": 1.5462,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9396686464520163,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00017410697117479823,
+      "loss": 1.7804,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9399812441387934,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017409047156931114,
+      "loss": 1.8893,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9402938418255705,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017407396749088787,
+      "loss": 1.5371,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9406064395123476,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017405745894052477,
+      "loss": 1.5866,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9409190371991247,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017404094591921853,
+      "loss": 1.5388,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9412316348859019,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017402442842796604,
+      "loss": 1.438,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9415442325726789,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017400790646776443,
+      "loss": 1.892,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9418568302594561,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017399138003961124,
+      "loss": 1.4763,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9421694279462332,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001739748491445041,
+      "loss": 1.6418,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9424820256330103,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017395831378344112,
+      "loss": 1.7746,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9427946233197875,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017394177395742047,
+      "loss": 1.8002,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9431072210065645,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017392522966744068,
+      "loss": 1.686,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9434198186933417,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017390868091450055,
+      "loss": 1.6964,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9437324163801187,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017389212769959922,
+      "loss": 1.656,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9440450140668959,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017387557002373596,
+      "loss": 1.6357,
+      "step": 3020
+    },
+    {
+      "epoch": 0.944357611753673,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017385900788791038,
+      "loss": 1.8136,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9446702094404501,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00017384244129312239,
+      "loss": 1.5841,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9449828071272273,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017382587024037212,
+      "loss": 1.5595,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9452954048140044,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017380929473066,
+      "loss": 1.6447,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9456080025007815,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017379271476498665,
+      "loss": 1.6323,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9459206001875586,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017377613034435315,
+      "loss": 1.62,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9462331978743357,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017375954146976058,
+      "loss": 1.4751,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9465457955611128,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017374294814221055,
+      "loss": 2.2368,
+      "step": 3028
+    },
+    {
+      "epoch": 0.94685839324789,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017372635036270472,
+      "loss": 1.7495,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9471709909346671,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001737097481322452,
+      "loss": 1.9299,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9474835886214442,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017369314145183426,
+      "loss": 1.5842,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9477961863082213,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017367653032247446,
+      "loss": 1.6439,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9481087839949984,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001736599147451686,
+      "loss": 1.7489,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9484213816817756,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017364329472091986,
+      "loss": 1.6981,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9487339793685526,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001736266702507316,
+      "loss": 1.98,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9490465770553298,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001736100413356074,
+      "loss": 1.5686,
+      "step": 3036
+    },
+    {
+      "epoch": 0.949359174742107,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017359340797655116,
+      "loss": 1.6756,
+      "step": 3037
+    },
+    {
+      "epoch": 0.949671772428884,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017357677017456715,
+      "loss": 1.6345,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9499843701156612,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017356012793065976,
+      "loss": 1.6958,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9502969678024382,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001735434812458337,
+      "loss": 1.6856,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9506095654892154,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017352683012109395,
+      "loss": 1.6888,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9509221631759925,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001735101745574458,
+      "loss": 1.7944,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9512347608627696,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001734935145558947,
+      "loss": 1.4633,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9515473585495468,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001734768501174465,
+      "loss": 1.5549,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9518599562363238,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017346018124310723,
+      "loss": 1.6942,
+      "step": 3045
+    },
+    {
+      "epoch": 0.952172553923101,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001734435079338832,
+      "loss": 1.8094,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9524851516098781,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017342683019078102,
+      "loss": 1.6422,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9527977492966552,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017341014801480748,
+      "loss": 1.4798,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9531103469834323,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001733934614069698,
+      "loss": 1.6282,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9534229446702095,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017337677036827534,
+      "loss": 1.5165,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9537355423569865,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017336007489973171,
+      "loss": 1.6635,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9540481400437637,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017334337500234687,
+      "loss": 1.7504,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9543607377305408,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017332667067712905,
+      "loss": 1.8412,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9546733354173179,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001733099619250867,
+      "loss": 1.616,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9549859331040951,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017329324874722847,
+      "loss": 1.7954,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9552985307908721,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017327653114456343,
+      "loss": 1.6591,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9556111284776493,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017325980911810085,
+      "loss": 1.6327,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9559237261644263,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017324308266885026,
+      "loss": 1.5621,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9562363238512035,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017322635179782138,
+      "loss": 2.0408,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9565489215379807,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017320961650602436,
+      "loss": 1.5293,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9568615192247577,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017319287679446949,
+      "loss": 1.5787,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9571741169115349,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001731761326641674,
+      "loss": 1.6182,
+      "step": 3062
+    },
+    {
+      "epoch": 0.957486714598312,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001731593841161289,
+      "loss": 1.6671,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9577993122850891,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017314263115136516,
+      "loss": 1.6618,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9581119099718662,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017312587377088756,
+      "loss": 1.6887,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9584245076586433,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017310911197570777,
+      "loss": 1.6217,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9587371053454204,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017309234576683778,
+      "loss": 1.7303,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9590497030321976,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001730755751452897,
+      "loss": 1.6497,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9593623007189747,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.000173058800112076,
+      "loss": 1.8203,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9596748984057518,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017304202066820948,
+      "loss": 2.1236,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9599874960925289,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001730252368147031,
+      "loss": 1.7534,
+      "step": 3071
+    },
+    {
+      "epoch": 0.960300093779306,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017300844855257008,
+      "loss": 1.6816,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9606126914660832,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.000172991655882824,
+      "loss": 1.5992,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9609252891528602,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017297485880647862,
+      "loss": 1.8889,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9612378868396374,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017295805732454804,
+      "loss": 1.6511,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9615504845264145,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017294125143804657,
+      "loss": 1.7686,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9618630822131916,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0001729244411479888,
+      "loss": 1.7564,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9621756798999688,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0001729076264553896,
+      "loss": 1.6458,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9624882775867458,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017289080736126409,
+      "loss": 1.698,
+      "step": 3079
+    },
+    {
+      "epoch": 0.962800875273523,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017287398386662764,
+      "loss": 1.684,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9631134729603001,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001728571559724959,
+      "loss": 1.7003,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9634260706470772,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017284032367988482,
+      "loss": 1.5827,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9637386683338544,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001728234869898106,
+      "loss": 1.7952,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9640512660206314,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017280664590328966,
+      "loss": 1.5528,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9643638637074086,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001727898004213387,
+      "loss": 1.8732,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9646764613941857,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017277295054497478,
+      "loss": 1.5453,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9649890590809628,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017275609627521508,
+      "loss": 1.8652,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9653016567677399,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017273923761307712,
+      "loss": 1.5761,
+      "step": 3088
+    },
+    {
+      "epoch": 0.965614254454517,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017272237455957868,
+      "loss": 1.3679,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9659268521412941,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00017270550711573788,
+      "loss": 2.1864,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9662394498280713,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001726886352825729,
+      "loss": 1.8203,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9665520475148484,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0001726717590611024,
+      "loss": 1.6397,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9668646452016255,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017265487845234524,
+      "loss": 1.7298,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9671772428884027,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017263799345732043,
+      "loss": 1.4412,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9674898405751797,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0001726211040770474,
+      "loss": 1.6235,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9678024382619569,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0001726042103125458,
+      "loss": 1.4866,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9681150359487339,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001725873121648355,
+      "loss": 1.8129,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9684276336355111,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00017257040963493663,
+      "loss": 1.7193,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9687402313222883,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00017255350272386968,
+      "loss": 1.6863,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9690528290090653,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00017253659143265534,
+      "loss": 1.5868,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9693654266958425,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017251967576231448,
+      "loss": 1.9038,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9696780243826195,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0001725027557138684,
+      "loss": 1.5963,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9699906220693967,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001724858312883386,
+      "loss": 1.9158,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9703032197561738,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001724689024867468,
+      "loss": 1.7879,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9706158174429509,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00017245196931011495,
+      "loss": 2.2104,
+      "step": 3105
+    },
+    {
+      "epoch": 0.970928415129728,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017243503175946542,
+      "loss": 1.3733,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9712410128165052,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001724180898358207,
+      "loss": 1.8072,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9715536105032823,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017240114354020368,
+      "loss": 1.6554,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9718662081900594,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0001723841928736373,
+      "loss": 1.8434,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9721788058768365,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017236723783714496,
+      "loss": 1.7078,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9724914035636136,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017235027843175027,
+      "loss": 1.4973,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9728040012503908,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017233331465847705,
+      "loss": 2.0236,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9731165989371678,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017231634651834946,
+      "loss": 1.55,
+      "step": 3113
+    },
+    {
+      "epoch": 0.973429196623945,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017229937401239188,
+      "loss": 1.5074,
+      "step": 3114
+    },
+    {
+      "epoch": 0.973741794310722,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017228239714162896,
+      "loss": 1.4308,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9740543919974992,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017226541590708566,
+      "loss": 1.8249,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9743669896842764,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017224843030978705,
+      "loss": 1.8337,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9746795873710534,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017223144035075864,
+      "loss": 1.7211,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9749921850578306,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017221444603102617,
+      "loss": 1.7391,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9753047827446076,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017219744735161554,
+      "loss": 2.0078,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9756173804313848,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.000172180444313553,
+      "loss": 1.7833,
+      "step": 3121
+    },
+    {
+      "epoch": 0.975929978118162,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017216343691786509,
+      "loss": 1.508,
+      "step": 3122
+    },
+    {
+      "epoch": 0.976242575804939,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001721464251655785,
+      "loss": 2.0652,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9765551734917162,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0001721294090577203,
+      "loss": 1.5267,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9768677711784933,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017211238859531774,
+      "loss": 1.838,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9771803688652704,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017209536377939846,
+      "loss": 1.7286,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9774929665520475,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0001720783346109901,
+      "loss": 1.8045,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9778055642388246,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001720613010911209,
+      "loss": 1.712,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9781181619256017,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001720442632208191,
+      "loss": 1.5521,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9784307596123789,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0001720272210011133,
+      "loss": 1.7718,
+      "step": 3130
+    },
+    {
+      "epoch": 0.978743357299156,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017201017443303242,
+      "loss": 1.4686,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9790559549859331,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00017199312351760555,
+      "loss": 1.6478,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9793685526727102,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00017197606825586204,
+      "loss": 1.4012,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9796811503594873,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017195900864883158,
+      "loss": 1.6166,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9799937480462645,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017194194469754407,
+      "loss": 1.7632,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9803063457330415,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00017192487640302969,
+      "loss": 1.497,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9806189434198187,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017190780376631886,
+      "loss": 1.756,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9809315411065959,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0001718907267884423,
+      "loss": 1.5489,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9812441387933729,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017187364547043091,
+      "loss": 1.5929,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9815567364801501,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.000171856559813316,
+      "loss": 1.7889,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9818693341669271,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017183946981812897,
+      "loss": 1.4263,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9821819318537043,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017182237548590162,
+      "loss": 1.8588,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9824945295404814,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017180527681766593,
+      "loss": 1.7062,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9828071272272585,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017178817381445418,
+      "loss": 1.5145,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9831197249140357,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001717710664772989,
+      "loss": 1.6806,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9834323226008127,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017175395480723286,
+      "loss": 1.9361,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9837449202875899,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00017173683880528917,
+      "loss": 1.5781,
+      "step": 3147
+    },
+    {
+      "epoch": 0.984057517974367,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017171971847250106,
+      "loss": 1.5337,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9843701156611441,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017170259380990216,
+      "loss": 1.8557,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9846827133479212,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017168546481852634,
+      "loss": 1.735,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9849953110346984,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017166833149940763,
+      "loss": 1.6696,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9853079087214754,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00017165119385358045,
+      "loss": 1.5103,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9856205064082526,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00017163405188207932,
+      "loss": 1.3137,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9859331040950297,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00017161690558593925,
+      "loss": 2.1945,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9862457017818068,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001715997549661953,
+      "loss": 1.8129,
+      "step": 3155
+    },
+    {
+      "epoch": 0.986558299468584,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017158260002388294,
+      "loss": 1.7308,
+      "step": 3156
+    },
+    {
+      "epoch": 0.986870897155361,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017156544076003778,
+      "loss": 1.7969,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9871834948421382,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017154827717569577,
+      "loss": 1.5541,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9874960925289152,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00017153110927189307,
+      "loss": 1.6279,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9878086902156924,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00017151393704966617,
+      "loss": 1.7777,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9881212879024696,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017149676051005176,
+      "loss": 1.7864,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9884338855892466,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001714795796540868,
+      "loss": 1.8507,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9887464832760238,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017146239448280853,
+      "loss": 1.5787,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9890590809628009,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017144520499725444,
+      "loss": 1.6532,
+      "step": 3164
+    },
+    {
+      "epoch": 0.989371678649578,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00017142801119846227,
+      "loss": 1.4543,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9896842763363551,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00017141081308747003,
+      "loss": 1.639,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9899968740231322,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00017139361066531605,
+      "loss": 1.6788,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9903094717099094,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017137640393303878,
+      "loss": 1.5768,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9906220693966865,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017135919289167707,
+      "loss": 1.6102,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9909346670834636,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017134197754226996,
+      "loss": 1.5106,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9912472647702407,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00017132475788585674,
+      "loss": 1.4294,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9915598624570178,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00017130753392347698,
+      "loss": 1.552,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9918724601437949,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00017129030565617053,
+      "loss": 1.4553,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9921850578305721,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00017127307308497752,
+      "loss": 1.6594,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9924976555173491,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001712558362109382,
+      "loss": 1.7315,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9928102532041263,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001712385950350933,
+      "loss": 1.5794,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9931228508909034,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0001712213495584836,
+      "loss": 1.7619,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9934354485776805,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00017120409978215034,
+      "loss": 1.6773,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9937480462644577,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00017118684570713476,
+      "loss": 1.5635,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9940606439512347,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017116958733447862,
+      "loss": 1.8061,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9943732416380119,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017115232466522379,
+      "loss": 1.496,
+      "step": 3181
+    },
+    {
+      "epoch": 0.994685839324789,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001711350577004125,
+      "loss": 1.9932,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9949984370115661,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00017111778644108707,
+      "loss": 1.7719,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9953110346983433,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017110051088829023,
+      "loss": 1.9202,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9956236323851203,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000171083231043065,
+      "loss": 1.7274,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9959362300718975,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017106594690645454,
+      "loss": 1.6006,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9962488277586746,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017104865847950224,
+      "loss": 1.8627,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9965614254454517,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00017103136576325194,
+      "loss": 1.6147,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9968740231322288,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017101406875874754,
+      "loss": 1.8255,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9971866208190059,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0001709967674670333,
+      "loss": 1.6937,
+      "step": 3190
+    },
+    {
+      "epoch": 0.997499218505783,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001709794618891538,
+      "loss": 1.7125,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9978118161925602,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0001709621520261537,
+      "loss": 1.7602,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9981244138793373,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017094483787907804,
+      "loss": 1.8293,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9984370115661144,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00017092751944897214,
+      "loss": 2.0362,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9987496092528916,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00017091019673688148,
+      "loss": 1.8003,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9990622069396686,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0001708928697438519,
+      "loss": 1.6969,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9993748046264458,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00017087553847092943,
+      "loss": 1.4631,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9996874023132228,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001708582029191604,
+      "loss": 2.0063,
+      "step": 3198
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00017084086308959132,
+      "loss": 1.4657,
+      "step": 3199
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 12796,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 3199,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.538493341728768e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}